Diseño Capacidad 6 PDF

PLASTIC HINGING BEHAVIOR OF REINFORCED CONCRETE BRIDGE
COLUMNS
BY
Zeynep Firat Alemdar
Submitted to the graduate degree program in Civil

Engineering and the Graduate Faculty of the
University of Kansas in partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
Chairperson
Committee Members*:
Date Defended: April 26, 2010
The Dissertation Committee for Zeynep Firat Alemdar certifies

that this is the approved version of the following dissertation:
PLASTIC HINGING BEHAVIOR OF REINFORCED CONCRETE BRIDGE

COLUMNS
Committee:
_________________________
Chairperson
_________________________
_________________________
__________________________
__________________________
Date approved: _________________________
ii
ABSTRACT
The location of inelastic deformations in reinforced concrete bridge columns has

been examined to simulate the nonlinear response of bridge columns and estimate the
ultimate displacement capacity. In bridge columns, these nonlinear deformations
generally occur over a finite hinge length. A model of hinging behavior in reinforced
concrete bridge columns will help guide proportioning, detailing and drift estimates for
performance-based design. Data was collected during the NEESR investigation of the
seismic performance of four-span large-scale bridge systems at the University of Nevada
Reno that details deformations in column hinging regions during response to strong
shaking events. In order to evaluate the plastic hinging regions, a photogrammetric
method was used to remotely track deformations of the concrete surface in the joint
regions. The surface deformations and rotations of a reinforced concrete bridge column
under dynamic loading has been examined and compared with the results obtained from
traditional instruments.
This research utilized the experimental data from photogrammetry measurements
of bridge column deformations to create a finite element model that realistically
represents hinging behavior in a reinforced concrete bridge pier. The three dimensional
finite element model of one column was defined with the cap beam on the top of the
column and the footing system under the circular column using ABAQUS Finite
Element software. The results of the FE model of the bridge column under dynamic
loading were obtained and compared with the photogrammetric measurements as well as
the data from the traditional instrumentations.
iii
Two plastic hinge length expressions for reinforced concrete bridge columns
under static and dynamic loadings have been developed by studying the available test
results in the literature. Many of the previous tests were conducted using the static
loading and for small-scale components. A few of the tests focused on bridge columns
and dynamic loading. Expressions that have been developed to estimate the plastic hinge
lengths have either been based on the maximum drift at the top of the column, or the
spread of plasticity in the hinging regions. An expression to calculate the maximum drift
capacity of a bridge column in double curvature has been derived by considering the
deformations due to flexure as influenced by the definition of plastic hinge length (lp),
and the bond-slip effect of the longitudinal reinforcement at the connections. Drift
capacity of a bridge column, which corresponds to a 20% reduction in lateral load
capacity on the descending branch of the response backbone curve, has been estimated
using the new expression and compared with the results that were obtained from the
earlier plastic hinge length expressions. The measured drift of the bridge column from
the four-span large-scale bridge system test was also compared with the calculated
responses from the expressions. The proposed equations provide the best estimate of
plastic hinge length for reinforced concrete bridge columns.
iv
ACKNOWLEDGEMENTS
I want to express my sincere gratitude to my advisors, Professors JoAnn

Browning and Adolfo Matamoros, for their commitment, guidance, support, and advice
throughout my research at the University of Kansas. I would like to thank Dr. Stan
Rolfe, Dr. Francis Thomas, and Dr. Daniel Stockli for serving on my dissertation
committee.
I would also like to thank for the effort made by the laboratory manager, Patrick
Laplace, and a master student, Robbie Nelson, at the University of Nevada Reno. I also
want to express gratitude to my colleague, Nick Hunt, for the countless discussions and
help about research.
This dissertation is dedicated to my family, especially to my husband, Fatih
Alemdar, for his great support and encouragement.
Zeynep Firat Alemdar
TABLE OF CONTENTS
LIST OF TABLES FOR CHAPTERS 1 AND 2.................................................................. iv

LIST OF FIGURES FOR CHAPTERS 1 AND 2 ................................................................ iv
LIST OF NOTATION ........................................................................................................xiii
1 INTRODUCTION............................................................................................................ 1
1.1
GENERAL............................................................................................................. 1
1.2
DETERMINATION OF PLASTIC HINGE LENGTH ........................................ 1
1.3
LITERATURE REVIEW OF PLASTIC HINGE LENGTH (STATIC TESTS).. 3
1.4
1.3.1
The Institution of Civil Engineers Committee Report (1962) ................... 4
1.3.2
ACI Limits (1968) ..................................................................................... 8
1.3.3
Park, Priestley, and Gill (1982) ................................................................. 9
1.3.4
Mander (1983) ......................................................................................... 11
1.3.5
Priestley and Park (1987)......................................................................... 13
1.3.6
Sakai and Sheikh (1989).......................................................................... 17
1.3.7
Tanaka and Park (1990)........................................................................... 17
1.3.8
Paulay and Priestley (1992) ..................................................................... 19
1.3.9
Soesianawati, Park and Priestley (1986), Watson and Park (1994) ........ 21
1.3.10
Sheikh and Khoury (1993), Sheikh, Shah and Khoury (1994)................ 24
1.3.11
Kovacic (1995) ........................................................................................ 27
1.3.12
Bayrak and Sheikh (1997, 1999) ............................................................. 28
1.3.13
Bae (2005) ............................................................................................... 31
1.3.14
Restrepo, Seible, Stephan, and Schoettler (2006).................................... 40
1.3.15
Phan V., Saiidi M.S., Anderson J., and Ghasemi H. (2007).................... 45
1.3.16
Berry, Lehman, and Lowes (2008) .......................................................... 46
LITERATURE REVIEW OF PLASTIC HINGE LENGTH (DYNAMIC
TESTS) ........................................................................................................................... 49
1.4.1
Dodd et al. (2000) .................................................................................... 49
1.4.2
Hachem et al. (2003)................................................................................ 52
1.5
SUMMARY......................................................................................................... 53
1.6
OBJECTIVE AND SCOPE................................................................................. 57
2 PHOTOGRAMMETRIC MEASUREMENTS OF CONCRETE COLUMN
vi
DEFORMATIONS......................................................................................................... 58
2.1
INTRODUCTION ............................................................................................... 58
2.2
PROOF-OF-CONCEPT TEST............................................................................ 59
2.3
LARGE SCALE FOUR-SPAN RC BRIDGE TEST .......................................... 67
2.4
2.3.1
Description of Specimen.......................................................................... 67
2.3.2
Experimental Setup.................................................................................. 69
2.3.3
Earthquake Loading................................................................................. 74
2.3.4
Results...................................................................................................... 76
2.3.4.1
Definition of Points on Surface ............................................................... 78
2.3.4.2
Displacements........................................................................................ 109
2.3.4.3
Rotations ................................................................................................ 122
CONCLUSION.................................................................................................. 130
3 MANUSCRIPT 1: PHOTOGRAMMETRIC MEASUREMENTS OF RC BRIDGE

COLUMN DEFORMATIONS .................................................................................... 133
4 MANUSCRIPT 2: MODELING SURFACE DEFORMATIONS AND HINGING
REGIONS IN REINFORCED CONCRETE BRIDGE COLUMNS........................... 159
5 MANUSCRIPT 3: PLASTIC HINGE LENGTH EXPRESSION FOR RC BRIDGE
COLUMNS................................................................................................................... 211
6 CONCLUSIONS .......................................................................................................... 235
REFERENCES FOR CHAPTERS 1 AND 2 .................................................................... 239
APPENDIX A.................................................................................................................... 247
APPENDIX B.................................................................................................................... 254
vii
iv
LIST OF TABLES FOR CHAPTERS 1 AND 2

Table 1.1 Details of column specimens and measured test results...11
Table 1.2 Experimental and predicted plastic hinge lengths.16
Table 1.3 Details of Column Specimens (Tanaka and Park 1990)...19

Table 1.4 Details of Column Specimens (Watson and Park 1994)... ...22
Table 1.5 Details of Specimens (Sheikh and Khoury 1993, 1994)...24
Table 1.6 Details of beams tested by Kovacic..27
Table 1.7 Details and test results of column specimens ...28
Table 1.8 Details of Test Specimens.30
Table 1.9 Comparisons of measured and proposed plastic hinge lengths.....38
Table 1.10 Measured Plastic Hinge Lengths.....44
Table 1.11 Details of the column properties.....48
LIST OF FIGURES FOR CHAPTERS 1 AND 2
Figure 1.1 Curvature and deflection relationships for a reinforced concrete cantilever
(Paulay and Priestley 1992). ................................................................................................3
iv
Figure 1.2 Dimensions, steel content, and steel strain locations (Ernst 1957). ...................7
Figure 1.3 Exact curvature distributions for deflection calculations (Priestley and Park
1987) ..................................................................................................................................13
Figure 1.4 Effects of Various Parameters on Plastic Hinge Lengths (Sakai and Sheikh
1989) ..................................................................................................................................17
Figure 1.5 Theoretical curvature relationships for a prismatic reinforced concrete

cantilever column (Paulay and Priestley 1992) .................................................................21
Figure 1.7 Relationship between Plastic Hinge Length and Shear Span-to-Depth Ratio
(Bae 2005) .........................................................................................................................35
Figure 1.8 Effect of Axial Load on Curvature and Compressive Strain Profiles (Bae
2005) ..................................................................................................................................36
Figure 1.9 Relationship between Plastic Hinge Length and Axial Load (Bae 2005)........37
Figure 1.10 Effect of Amount of Longitudinal Reinforcement (Bae 2005) ......................38
Figure 1.11 Comparisons of Plastic Hinge Length (Eq. (1.20) versus Analysis) (Bae
2005) ..................................................................................................................................39
Figure 1.12 Idealization of curvature distribution in column: (a) column; (b) BMD; (c)
curvature diagram; and (d) equivalent curvature diagram (Restrepo et al. 2006) .............41
Figure 1.13 Plasticity spread coefficient ((Restrepo et al. 2006) ...................................43

v
vi
Figure 1.14 Strain penetration coefficient (Restrepo et al. 2006)...................................44
Figure 1.15 Equivalent plastic hinge length as ratio of column diameter (Restrepo et al.
2006) ..................................................................................................................................44
Fig. 1.16 Simulated and observed force-drift ratio for columns with different design
parameters (Berry et al. 2008) ...........................................................................................49
Fig. 2.1 Cross-section of the column .................................................................................60
Fig. 2.2 Grid Setup.............................................................................................................61
Fig. 2.3 Location of the column and the tower position....................................................62
Fig. 2.4 The aluminum tower setup ...................................................................................63
Fig. 2.5 Acceleration history of the Rinaldi earthquake record.........................................64
Fig. 2.6 Displacement @ Grid Level A vs. Time (Rinaldi 0.95g) (1 in. = 254 mm) ........66
Fig. 2.7 Rotation @ Grid Level A vs. Time (Rinaldi 0.95g).............................................66
Fig. 2.8 Elevation view of the four-span bridge ................................................................68
Fig. 2.9 Elevation and side view of the Bent1 ...................................................................68
Fig. 2.10 Elevation and side view of the Bent2 .................................................................69
Fig. 2.11 Elevation and side view of the Bent 3 ................................................................69

vi
vii
Fig. 2.12 Grid systems on the Bent3 east column surface.................................................71
Fig. 2.13 Close view of (a) bottom and (b) top grid systems ............................................71
Fig. 2.14 The aluminum tower and four cameras ..............................................................72
Fig. 2.15 Metal weight pieces placed on each side of the tower .......................................72
Fig. 2.16 Location of the Bent 3 east column and the aluminum tower............................73
Fig. 2.17 Grid system and LVDT locations on column in the Bent 3 east column...........74
Fig. 2.18 Fix point on the wall at the back of the column .................................................77
Fig. 2.19 Lateral displacement of fixed point on the wall at Test 4D (1 in. = 254 mm) ...78
Fig. 2.20 Vertical displacement of fixed point on the wall at Test 4D (1 in. = 254 mm)..78
Fig. 2.21 Point 7 vertical displacement at the bottom grid system (1 in. = 254 mm)........82
Fig. 2.22 Five second interval to compare even- odd- and combination-line analyses (1
in. = 254 mm) ....................................................................................................................83
Fig. 2.23 Comparison of Point 7 vertical displacement for even lines with LVDT 3EBR7
data (1 in. = 254 mm) ........................................................................................................84
Fig. 2.24 Comparison of Point 7 vertical displacement for odd lines with LVDT 3EBR7
data (1 in. = 254 mm) ........................................................................................................85
vii
viii
Fig. 2.25 LVDT 3EBR7 vertical displacement history (1 in. = 254 mm) .........................85
Fig. 2.26 Rotation of horizontal line calculated from Points 7 and 8 at the bottom grid in
the transverse direction ......................................................................................................86
Fig. 2.27 Close up of lines used to define Point 3 (Fig. 2.17) ...........................................87
Fig. 2.28 Pixel intensities of two different levels in an image ..........................................88
Fig. 2.29 Comparison of Point 7 vertical displacement with LVDT 3EBR7 data (1 in. =
254 mm).............................................................................................................................89
254 mm).............................................................................................................................89
Fig. 2.31 Rotation of horizontal line using Points 7 and 9 ................................................90
Fig. 2.32 Average rotation of four Points on the column surface......................................91
Fig. 2.33 Four corners surrounding general location of Point 12 and 13 ..........................92
Fig. 2.34 Corner (a) rotation of Point 12 and 13 ...............................................................93
Fig. 2.35 Corner (b) rotation of Point 12 and 13 ...............................................................93
Fig. 2.36 Corner (c) rotation of Point 12 and 13 ...............................................................94
Fig. 2.37 Corner (d) rotation of Point 12 and 13 ...............................................................94
viii
ix
Fig. 2.38 Average rotation of Point 12 and 13 ..................................................................95
Fig. 2.39 Point 7 vertical displacement vs. LVDT 3EBR7 data (1 in. = 254 mm)............97
Fig. 2.40 Point 9 vertical displacement vs. LVDT 3EBR8 data (1 in. = 254 mm)............97
Fig. 2.41 Rotation of horizontal line obtained using the constant Robert threshold .........98
Fig. 2.42 Point 7 vertical displacement compared with LVDT 3EBR7 data (1 in. = 254
mm)....................................................................................................................................98
mm)....................................................................................................................................99
Fig. 2.44 Rotation of horizontal line at h=7.7 in. (196 mm) from the bottom fixity.........99
254 mm)...........................................................................................................................101
254 mm)...........................................................................................................................101
Fig. 2.47 Rotation of horizontal line calculated using long-line .....................................102
Fig. 2.48 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data
using
the maximum intensity approach.....................................................................................103
ix
Fig. 2.49 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data using
the Edge Lines method (1 in. = 254 mm) ........................................................................104
Fig. 2.50 Rotation of vertical line calculated using Point 3 and 8...................................106
Fig. 2.51 Rotation of horizontal line using Point 7 and 9................................................106
Fig. 2.52 Comparison of rotations (local vs. average) in Test 4D...................................107
Fig. 2.53 Rotation of vertical line calculated using Point 38 and 52...............................107
Fig. 2.54 Rotation of horizontal line using Point 44 and 46............................................108
Fig. 2.55 Comparison of rotations (local vs. average) in Test 4D...................................108
Fig. 2.56 Comparison of Drift at Point 59 for Test 2 (1 in. = 254 mm) ..........................110
Fig. 2.57 Comparison of Drift at Point 59 for Test 4D (1 in. = 254 mm) .......................111
Fig. 2.58 Comparison of Drift at Point 59 for Test 6 (1 in. = 254 mm) ..........................111
Fig. 2.59 Maximum lateral movement ratios between photogrammetry and LVDT results
. ........................................................................................................................................112
Fig. 2.60 Overall deformed shape along the column height at maximum column drift
. ........................................................................................................................................114
Fig. 2.61 Top grid deformed shape with picture comparison at maximum column drift 115
xi
Fig. 2.63 The overall deformed shape along the column height at maximum column drift
. ........................................................................................................................................117
Fig. 2.64 Top grid deformed shape with picture comparison at maximum column drift
. ........................................................................................................................................118
Fig. 2.65 Bottom grid deformed shape with picture comparison at maximum column drift
. ........................................................................................................................................118
. ........................................................................................................................................120
Fig. 2.69 Rotation of vertical line calculated using Point 3 and 8 for Test 2 ..................123
Fig. 2.70 Rotation of vertical line calculated using Point 8 and 13 for Test 2 ................123
Fig. 2.71 - Rotation of vertical line calculated using Point 38 and 45 for Test 2............124
Fig. 2.72 - Rotation of vertical line calculated using Point 45 and 52 for Test 2............124
Fig. 2.73 - Rotation of vertical line calculated using Point 3 and 8 for Test 4D .............125
Fig. 2.74 Rotation of vertical line calculated using Point 8 and 13 for Test 4D .............126
Fig. 2.75 Rotation of vertical line calculated using Point 38 and 45 for Test4D ............126
Fig. 2.76 Rotation of vertical line calculated using Point 45 and 52 for Test4D ............127
Fig. 2.77 Rotation of vertical line calculated using Point 3 and 8 for Test 6 ..................127
xi
xii
Fig. 2.78 Rotation of vertical line calculated using Point 8 and 13 for Test 6 ................128
Fig. 2.79 Rotation of vertical line calculated using Point 38 and 45 for Test 6 ..............128
Fig. 2.80 Rotation of vertical line calculated using Point 45 and 52 for Test 6 ..............129
Fig. 2.81 Maximum rotation ratios between photogrammetry compared LVDT results
. ........................................................................................................................................130
xii
xiii
LIST OF NOTATION
p : Plastic curvature
m : The maximum curvature
y : Curvature at first yield
u : The ultimate curvature
lp : The plastic hinge length
l : The height of a column
p : The plastic displacement
d : Distance from extreme compression fiber to centroid of tension reinforcement, in.
(mm)
w : Uniformly distributed load at a section of maximum moment kip/in., (kN/mm)
Vz : Shear adjacent to a concentrated load or reaction at a section of maximum moment
kips, (kN)
Mm : Maximum moment in a length of member kip-in., (kN.mm)
Me : Elastic-limit resisting moment kip-in., (kN.mm)
Mu : Ultimate resisting moment concurrent with Pu kip-in., (kN.mm)
Pu : Ultimate resisting axial load kips, (kN)
cue : Elastic component of cu, either calculated or assumed in the range 0.001 to 0.002,
cu : Maximum compressive strain in concrete at Mu and Pu, and
cuo : Basic maximum compressive strain in concrete
p : The plastic rotation,
k1 : The influence of the type of reinforcing steel,
k2 : The influence of column load (when present)
k3 : The influence of strength of concrete
xiii
xiv
z : The distance of critical section to point of contraflexure

P : The column load acting in conjunction with bending moment,
Pu : The load capacity as an axially loaded column
cu : The cube strength of concrete.
db : The longitudinal bar diameter
Pe : The design compressive load of the column due to gravity and seismic loading
: Strength reduction factor
Ag : Gross area of column section, in2, (mm2).
Ach : Area of core concrete measured out-to-out of transverse reinforcement
fyt : Yield strength of transverse reinforcement
s : Spacing of transverse reinforcement
bc : Cross sectional dimension of column core, measured center-to-center of transverse
reinforcement
xiv
INTRODUCTION
Equation Chapter 1 Section 1

1.1
GENERAL
The determination of magnitude and location of inelastic deformations in
reinforced concrete bridge columns is a critical step for characterizing the performance
of the bridge system in earthquake events. Although it is possible to believe that some
ductility will be provided by beam hinges in bridge systems, it is generally the columns
of the bridges that must have inelastic rotational capacity. Bridge systems are designed to
keep inelastic behavior in the columns and away from the superstructure, which is
different than building systems.
If inelastic flexural deformations occur in a reinforced concrete structure due to
gravity and lateral loads, these deformations can generally be considered as concentrated
over a finite hinge length. The spread of plasticity, or hinge length is an important factor
in the analysis of deformation in reinforced concrete bridge structures and includes
elastic, plastic and softening stages of response. Little work has been completed to
determine hinge lengths in the plastic and softening phases of full-scale reinforced
concrete bridge systems. Previous work has relied on determining hinge lengths for
small-scale component tests. An evaluation of long-scale system behavior, as described
in this study, has the benefit of including the effects of moment redistribution and
progression of yielding throughout the entire structure.
1.2
DETERMINATION OF PLASTIC HINGE LENGTH

Plastic hinges occur in the sections that have bending moments that exceed the
nominal bending moment associated with yielding of the section. The equivalent plastic
hinge length can be calculated based on integration of the curvature distribution for
typical members. To simplify the calculations, an equivalent plastic hinge length lp can
be defined over which the plastic curvature, p, is assumed equal to m- y, where m is
the maximum curvature and y is the yield curvature, as shown in Fig. 1.1(a). The length
lp is determined so that the plastic displacement at the top of the cantilever column, p,
predicted from a displacement design method or from an experiment is the same as that
obtained from the actual curvature distribution as shown in Fig. 1.1(a). The lumped
plastic rotation, p, along the plastic hinge length is then computed as Eq. 1.1:
p = pl p = (m y ) l p
(1.1)
The plastic rotation derived using Eq. (1.1) can be used to determine the
displacement capacity of a section that experiences inelastic deformations. If the plastic
rotation is assumed to be concentrated at midheight of the plastic hinge, the plastic
displacement at the top of the cantilever column then becomes Eq. (1.2):
p = p (l 0.5l p ) = ( m y ) l p (l 0.5l p )
(1.2)
where l is the height of the column. The maximum nonlinear drift is then obtained from
the plastic displacement at the top of the cantilever. Therefore, a consistent prediction of
a plastic hinge length is necessary to examine the theoretical drift capacity of bridge
columns.
The hinge length indicates the theoretical length of damage concentration along
the column. Although the plastic hinge length should not be considered the required
column confinement region, it does indicate the minimum theoretical dimension. The
actual confinement region, which was shown as the extent of plasticity in Fig. 1.1 (a),
should be longer than the plastic hinge length.
y p
Extent of
plasticity
lp
(a) Curvature at maximum response
(b) Deflections
Figure 1.1 Curvature and deflection relationships for a reinforced concrete cantilever
(Paulay and Priestley 1992).
1.3
LITERATURE REVIEW OF PLASTIC HINGE LENGTH (STATIC TESTS)

The plastic hinge length from static tests has been studied widely by many
researchers (Baker 1956, Baker and Amarakone 1964, Mattock 1964, 1967, Corley 1966,
Park, Priestley, and Gill 1982, Priestley and Park 1987, Paulay and Priestley 1992,
Sheikh and Khoury 1993, Mendis 2001). A detailed review of formulae that are available
to calculate the plastic hinge length for reinforced concrete columns are presented in this
section. Few studies, however, have been conducted to determine the plastic hinge length
in reinforced concrete bridge columns. These studies include Park, Priestley, and Gill
1982, Priestley and Park 1987, Tanaka and Park 1990, Watson and Park 1994, Kovacic
1995, Dodd et al. 2000, Hachem et al. 2003, Bae 2005, and Phan 2007, and are also
included in this section. In the dynamic tests, the equation proposed by Priestley et al.
(1992) was used to estimate the plastic hinge length before testing the bridge columns,
even though this equation was derived using the static test results of columns. Two
studies have been done to determine the plastic hinge length with dynamic testing of
3
reinforced concrete bridge column under dynamic base excitation (Dodd et al. 2000, and
Hachem et al. 2003) and these will be discussed later in this chapter.
1.3.1
The Institution of Civil Engineers Committee Report (1962)
The Institute of Civil Engineers committee published a report on the ultimate load
design of concrete structures (I.C.E. 1962), which includes the principles of ultimate load
theory and its application to design.
The report specifies a conservative empirical
method of calculating the equivalent plastic hinge length lp,

lp
z
= k1k 2 k 3 ( ) 0.25
d
d
(1.3)
where k1, k2, and k3 represents the influence of the type of reinforcing steel, column load
(when present) and strength of concrete in lp respectively, z is the distance of critical
section to point of contraflexure, and d is the effective depth. The coefficients k1, k2, and
k3 were determined by examining several series of test results as described next.
From the analysis of test results that are described in this section (Ernst 1957,
McCollister et al. 1954, Poologasoundranayagam 1960, and Chan 1955), conservative
limiting values for k1, k2, and k3 are given as
0.7
k1 =
0.9
k2 = 1+0.5
0.6
k3 =
0.9
mild steel
cold worked steel

P
Pu
(1.4)
(1.5)
cu = 6, 000 lb / in 2 (42 MPa )
cu = 2, 000 lb / in 2 (14 MPa )
(1.6)
where P is the column load acting in conjunction with bending moment, Pu is the load
capacity as an axially loaded column, and cu is the cube strength of concrete.
Ernst (1957) tested 30 simply-supported beams having 6 in. (150 mm) by 12 in.
(305 mm) cross-section, 10 ft. (3050 mm) in length, and a span of 9 ft. (2750 mm) under
central point loading. The main parameters were widths of column stubs and tension
steel as shown in Fig. 1.2. Column stubs changed in width from 6 to 36 in. (150 mm to
915 mm) at mid-span. The range of tension steel reinforcement ratios was from 1% to
5%. The nominal 28-day cube strength of the concrete was 3000 psi (22 MPa) or 4000
psi (28 MPa). Steel yield strength was approximately 45 ksi (310 MPa).
McCollister et al. (1954) designed 31 beams with 6 in. (150 mm) by 12 in. (305
mm) cross-section and 9 ft. (2750 mm) span. The main variables comprised the concrete
strength from 1905 psi (13 MPa) to 6407 psi (44 MPa), tension and compression steel
content (from 0.17% to 5.10% and from 0% to 4.08%, respectively), and the effect of
column stubs with dimensions of 6 in. (150 mm) by 6 in. (150 mm) cross-section and 12
in. (305 mm) height.
Poologasoundranayagam (1960) tested 38 simply supported beams having 4 in.
(100 mm) by 9 in. (230 mm) cross section and spans of 4-6 ft. (1220-1830 mm) under
central point loading. The principal factors were strength of concrete from 2,385 psi (16
MPa) to 6,330 psi (44 MPa), percentage of tension reinforcement (0.62% to 5.1%), and
quality of steel (mild or cold-worked and post-tensioned high tensile wire).
Chan (1955) conducted three series of tests. The first series consisted of nine short
prisms having 6 in. (150 mm) square sections and 11 in. (290 mm) long. They were
reinforced with 4 5/8 in. diameter (16 mm) longitudinal bars and used ties. These prisms
were pin-ended and loaded under compressive load with an eccentricity of in. (13
mm). Seven cylinders of 6 in. (150 mm) diameter and 12 in. (305 mm) length were
tested in the second series. They had the same longitudinal reinforcement with the first
series, but they were loaded in compression with an eccentricity of 1/4 in. (7 mm). The
main variable was lateral binding, where spiral reinforcement of in. (7 mm), 3/16 in.
(5 mm), and 1/8 in. (3 mm) diameter bars and pitches from 1 in. (25 mm) to 4 in. (100
mm) were used to confine the specimens. The last series included seven struts with 6 in.
(105 mm) by 3-5/8 in. (90 mm) cross section and 52 in. (1320) long. They were
symmetrically reinforced with 4 1/2 in. diameter (13 mm) bars, laterally bound with
rectangular welded closed links spaced at 3 in. (75 mm). They were tested under axial
compression through pins at the ends, and a central lateral point load. It was intended to
simulate a plastic hinge formation within the critical region of a column under bending
moment and high axial load. The nominal 28-day cube strength of the concrete was 3000
psi (22 MPa) for the first and second series, and 4000 psi (28 MPa) for the last series.
Average yield strength of the steel was around 40 ksi (275 MPa).
Chan (1962) evaluated the methods and parameters that were recommended in
the ICC report (1962) by studying thirteen column tests, comprising six by
Poologasoundranayagam (1960) and seven by Chan (1955), covering a range of P/Pu
from 0.06 to 0.78, cube strengths from 2,380 (16 MPa) to 5,160 psi (36 MPa), and
symmetrically reinforced steel ratios from 1.23% to 1.92%. Tests were analyzed and
compared in order to calculate values of EIe and p (plastic rotation) as described below.
Figure 1.2 Dimensions, steel content, and steel strain locations (Ernst 1957).
The testing procedure was similar to the study of 54 beam tests done by Chan
(1954); in addition, values of k2 were included to consider the effect of the column
loading. The histograms of the column test results were expressed as the ratio of
experimental to calculated values of EIe/ EIe and p/ p. They had a similar distribution
with the beam test results. Chan observed that the parameters used in Eq. (1.3) were safe
and statistically acceptable, however, the number of test results was small and more tests
were desired. The author also reported that in a broad range of structural members lp
varies from about 0.4 to 2.4d.
1.3.2
ACI Limits (1968)
The ACI-ASCE Committee 428 on Limit Design (1968) recommended upper and
lower plastic hinge limits rather than a single equation. The length along a member from
the section of maximum moment, lp, should be bigger than the lesser of the two values
given in Eq. (1.7) and the value given in Eq. (1.8):
d
R + 0.03 zRm
4
and
Rd
but not greater than R + 0.10 zRm

2
(1.7)
(1.8)
in which;
R =
z=
0.004 cue
,
cuo cue
(1.9)
4M m
(1.10)
4V z + wM m Rm
Rm =
Mm Me
Mu Me
(1.11)
where;
d = distance from extreme compression fiber to centroid of tension reinforcement, in.

(mm)
w = uniformly distributed load at a section of maximum moment kip/in., (kN/mm)
Vz = shear adjacent to a concentrated load or reaction at a section of maximum moment
kips, (kN)
Mm = maximum moment in a length of member kip-in., (kN.mm)
Me = elastic-limit resisting moment kip-in., (kN.mm)
Mu = ultimate resisting moment concurrent with Pu kip-in., (kN.mm)
Pu = ultimate resisting axial load kips, (kN)
cue = elastic component of cu, either calculated or assumed in the range 0.001 to 0.002,
cu = maximum compressive strain in concrete at Mu and Pu, and
cuo = basic maximum compressive strain in concrete (neglecting possible amplifying
influences of confinement, loading rate and strain gradients) to which a value in the
range 0.003 to 0.004 needs to be assigned.
R restricts the range of total inelastic rotation by providing reduced limits on
hinge lengths for the greater assumed values of inelastic strains and curvatures, and
increased limits on hinge lengths for the smaller assumed strain values [ACI-ASCE
(1968)].
The formulae suggested by ACI 428 committee can be utilized as lower and
upper limits for inelastic analysis of normal and high-strength concrete structures. The
ACI formulae, however, do not rely on longitudinal and lateral reinforcement ratios.
1.3.3
Park, Priestley, and Gill (1982)

Four full-size reinforced concrete columns with 22 in. (550 mm) square sections
and 10.8 ft. (3300 mm) in height were tested by Park et al. (1982). The longitudinal
9
reinforcement in each column consisted of twelve 0.94-in. (24-mm.) diameter deformed

bars having a reinforcement ratio of 1.79%. The yield strength of the longitudinal steel
was 55.1 ksi (380 MPa). The transverse steel was plain round bars and the yield strength
was 40 ksi (275 MPa). The ranges of the applied axial loads were from 0.2fcAg to
0.6fcAg. Details of column specimens are given in Table 1.1.
The equivalent plastic hinge length, lp, was calculated by using the Eq. (1.2) for
the last load cycle in the test. The plastic displacement, p, was measured beyond the
first yield displacement and plastic curvature, u- y, was measured beyond the first yield
curvature over the 3.9 in. (100 mm) gage length adjacent to the central stub, where u is
the ultimate curvature and y is the yield curvature.
Based on the tests of the four reinforced concrete columns, Park et al. (1982)
showed that the calculated equivalent plastic hinge lengths were insensitive to axial load
level and had an average value of 0.42h, where h is the overall section depth. Table 1.1
lists the calculated plastic hinge length results for the tested columns and the ratio of this
length to the section depth.
Based on the limited column tests in this study, Park et al. concluded that lp =
0.4h can be used as a simple and safe approximation for plastic hinge lengths in columns.
It should be noted that lp is the equivalent length of plastic hinge to be used in evaluating
the ultimate curvature requirements, and should not define the length of the column that
needs to be confined along the critical section.
10
Table 1.1 Details of column specimens and measured test results
Axial Load
Longitudinal
Reinf.
fc,
ksi
(Mpa)
(1)
3.35
(23)
P,
fy,
kips
ksi
l,
P/fcAg (Mpa)
%
(kN)
(3)
Unit
(5)
(2)
(4)
408.03
55.1
1
0.26
1.79
(1815)
(380)
602.49
55.1
2
6 (41)
0.214
1.79
(2680)
(380)
3.1
611.25
55.1
3
0.42
1.79
(21)
(2719)
(380)
3.41 958.81
55.1
4
0.6
1.79
(24)
(4265)
(380)
1: Compressive cylinder strength of concrete
Transverse Reinf.
db,
in.
(mm)
(6)
0.39
(10)
0.47
(12)
0.39
(10)
0.47
(12)
fyh,
ksi
(Mpa)
(7)
43.07
(297)
45.83
(316)
43.07
(297)
42.64
(294)
s,
%
(8)
Ash/Ash,ACI
(9)
1.5
0.66
2.3
0.63
0.89
3.5
1.47
Meas.
Lp,
in.
(mm)
(10)
9.53
(242)
7.44
(189)
8.62
(219)
10.75
(273)
Lp/h
(11)
0.44
0.34
0.4
0.5
2: Applied axial load

3: Axial load ratio
4: The yield strength of longitudinal steel
5: The longitudinal reinforcement ratio
6: The diameter of the longitudinal reinforcement
7: The yield strength of transverse steel
8: The volumetric ratio of transverse reinforcement to core concrete
9: The ratio of total effective area of rectangular hoop bars to that required by ACI
10: Measured plastic hinge length
11: The ratio of measured plastic hinge length to the depth of the column
1.3.4
Mander (1983)
Experimental studies conducted by previous investigators at the University of
Canterbury (Gill et al. 1979, Potangaroa et al. 1979, Ghee et al. 1981, Davey et al. 1975,
Munro et al. 1976 and Heng et al. 1978) have supported the theory that the equivalent
plastic hinge length, lp, may vary from 0.35 to 0.65 of the overall member depth for solid
reinforced concrete columns. Based on a comparison of the available results for
octagonal specimens (RRU 1983), it was found that the equivalent plastic hinge length is
11
independent of the axial load level and aspect ratio. A value of lp = 0.5D was
recommended.
After examining the experimental results studied at the University of Canterbury,
Mander concluded that contributions to plastic deformation were primarily from two
sources: (i) the spread of plasticity along the member length due to the moment gradient
and (ii) yield penetration of the longitudinal reinforcement beyond the limits of the
plastic hinge. The equivalent length of yield penetration, Lpy, could be written in terms of
the longitudinal bar diameter from the force-deflection analyses:
L py = 6.35 d b
(in.)
(1.12a)
L py = 32 d b
(mm)
(1.12b)
where db is the longitudinal bar diameter.

The additional plastic hinge length due to the spread of plasticity along the
member length was found to be approximately six percent of the column length, L, after
analyzing all the test results. Thus, the equivalent plastic hinge length can be calculated
from the equation below.
L p = L py +0.06 L
(1.13)
When the predicted and observed results are compared, Eq. (1.13) generally
provides a conservative prediction of the equivalent plastic hinge length. Mander also
noted that Eq. (1.13) must not be used for estimating the length requiring detailed
confinement because plastic curvature would spread over approximately three equivalent
plastic hinge lengths.
12
1.3.5
Priestley and Park (1987)

Instead of obtaining the plastic hinge length using a linear elastic curvature
distribution along the column, an alternate approach was developed by Priestley and Park
(1987) considering the moment-curvature relationships for different sections along the
height of the column. The curvature distribution along the column can be calculated
using Eq. (1.14) for any given base moment as shown in Fig. 1.3(a). The predicted
displacement at the top of the column is then obtained by integrating the curvature
profile.
L
= ( x ) xdx
(1.14)
Using an incremental analysis based on this procedure results in theoretical

difficulties when the moment-curvature relationship has a curve with strength
degradation (negative slope). Failure is predicted when the column reaches the maximum
load. Sections having moment demands that are less than their capacity are assumed to
keep their prior curvatures past the post-peak load behavior.
(a)
(b)
Figure 1.3 Exact curvature distributions for deflection calculations (Priestley and Park
1987)
13
Priestley and Park (1987) reported that an elasto-plastic approximation should

consider a plastic hinge length proportional to the column height L, because the
predicted curvature distribution for columns that have identical section dimensions but
different heights would be geometrically similar. Although this relationship between the
column height and the plastic hinge length was accepted by early models for plastic
hinge length (Baker 1964; Corley 1966), it was not supported by previous experimental
observations because of two reasons as explained below:
1. The first reason is the slip of longitudinal reinforcement relative to the
concrete. Within the plastic hinge region, slip of reinforcement also leads
to longitudinal reinforcement strains at sections above the base to be
higher than expected. Therefore, the length of yield penetration and
resultant slippage will definitely be independent of column height L, and
would depend mainly on the diameter of the longitudinal reinforcement.
2. The second reason is the influence of shear on the crack pattern. If
flexural cracks are inclined under the influence of shear, the planesections-remain-plane hypothesis will not be valid, and steel stress and
strain will increase above the levels estimated based on the plane-sections
hypothesis. This leads to a spread the plasticity, and increases the plastic
hinge length. Figure 1.2(b) illustrates the effect of yield penetration and
spreading of plasticity due to shear. The lateral deflection of the center of
mass of the column is calculated by integrating the modified curvature
distribution.
Based on the arguments above, concrete bridge column tests were conducted in
two stages. In the first stage, sections including square, rectangular, and circular shapes
14
were tested under axial load only. The range of longitudinal reinforcement ratios was
between 1% and 4%, and the lateral reinforcement ratio was from 0.5% to 1.5% with
spiral or circular hoops. The axial load values ranged from 0.2Po to 0.7Po, where
Po=Pe/fcAg (Pe is design compressive load of the column due to gravity and seismic
loading, fc is compressive cylinder strength of concrete, and Ag is gross area of section).
In the second stage, the sections included square, diagonal, octagonal, and hollow square
shapes were tested under continued axial load and cyclic reversals of bending moment.
The test columns were instrumented extensively along the potential plastic hinge regions.
Priestley and Park (1987) proposed a general plastic hinge length formula (Eq.
1.15) based on the new test results.
l p = C1 L + C 2 d b + C 3 D
(1.15)
where L is the distance from the point of contraflexure of the column to the section of
maximum moment, db is the longitudinal bar diameter, D is section depth (or diameter
for circular sections) and C1, C2, and C3 are constants determined from curvature
distributions along the column length of the specimen.
Curvature distributions along the length were obtained for all units during the column
tests to predict the values of the constants. Best fit values of C1 = 0.08, C2 = 6, and C3 =
0 were found based on the analysis of the test results. Therefore, Eq. 1.15 becomes
l p = 0.08 L + 6d b
(ksi)
(1.16a)
l p = 0.08 L + 0.88d b
(MPa)
(1.16b)
Priestley and Park (1987) obtained good agreement between the experimentally
derived values for lp and values calculated using Eq. (1.16). The tests that were evaluated
included studies outside this program (Gill et al. 1982, Potangaroa et al. 1981, Davey et
al. 1975, Munro et al. 1976, Ng et al. 1978, Ghee et al. 1985 and Mander et al. 1984).
15
The average hinge length that was calculated for all tests was approximately equal to lp=
0.5D as shown in Table 1.2. The experimental data did not show any relationship
between plastic hinge length and axial load ratio, longitudinal reinforcement ratio, or
yield stress of longitudinal reinforcement.
Table 1.2.Experimental and predicted plastic hinge lengths
Researchers
Davey et al.
Munro et al.
Heng et al.
Gill et al.
Potangaroa et al.
Ghee et al.
Mander et al.
Column
aspect ratio
4
4
5.5
4
2.18
2.18
2.18
2.18
2
2
2
2
4
4
4
4
4.27
4.27
4.27
4.27
Plastic hinge length Lp

Section width, Longitudinal bar
D or h, in.
diameter, db, in.
Predicted Experiment/
(mm)
(mm)
Experiment
(Eq. 1.13a) Predicted
D=19.69 (500)
0.51 (13)
0.54D
0.44D
1.23
D=19.69 (500)
0.51 (13)
0.58D
0.60D
0.97
D=19.69 (500)
0.51 (13)
0.45D
0.60D
0.76
D=9.84 (250)
0.51 (13)
0.58D
0.64D
0.91
h=21.65 (550)
0.94 (24)
0.44h
0.44h
1
h=21.65 (550)
0.94 (24)
0.34h
0.44h
0.77
h=21.65 (550)
0.94 (24)
0.40h
0.44h
0.91
h=21.65 (550)
0.94 (24)
0.50h
0.44h
1.13
D=23.62 (600)
0.94 (24)
0.35D
0.40D
0.88
D=23.62 (600)
0.94 (24)
0.35D
0.40D
0.88
D=23.62 (600)
0.94 (24)
0.37D
0.40D
0.93
D=23.62 (600)
0.94 (24)
0.42D
0.40D
1.05
D=15.75 (400)
0.63 (16)
0.54D
0.56D
0.96
D=15.75 (400)
0.63 (16)
0.61D
0.56D
1.09
h=15.75 (400)
0.63 (16)
0.73h
0.56h
1.3
h=15.75 (400)
0.63 (16)
0.55h
0.56h
0.98
h=29.53 (750)
0.39 (10)
0.37h
0.42h
0.88
h=29.53 (750)
0.39 (10)
0.38h
0.42h
0.9
h=29.53 (750)
0.39 (10)
0.40h
0.42h
0.95
h=29.53 (750)
0.39 (10)
0.41h
0.42h
0.98
Average =
0.97
16
1.3.6
Sakai and Sheikh (1989)

Based on a review of the literature, Sakai and Sheikh (1989) noted that the plastic
hinge length increased as the aspect ratio (L/h or L/D, which is equivalent to the shear
span-to-depth ratio for cantilever columns) increased. Bilinear curves were developed to
give the relationship between the plastic hinge length and the aspect ratio as shown in
Figure 1.4. They concluded that the amount of transverse reinforcement, axial load level,
and aspect ratio had an influence on the plastic hinge length. The plastic hinge length
generally increased with increasing values of each parameter.
Figure 1.4 Effects of Various Parameters on Plastic Hinge Lengths (Sakai and Sheikh
1989)
1.3.7
Tanaka and Park (1990)

Tanaka and Park (1990) completed two series of column tests. In the first series,
four column specimens (Units 1 to 4) had a total height of 5.9 ft. (1800 mm) and 16 in.
17
(400 mm) square sections. In the second series, four column specimens (Units 5 to 8) had
a total height of 5.4 ft. (1650 mm) and 22 in. (550 mm) by 22 in. (550 mm) square
sections. The shear span-to-depth ratios were 4 and 3 for the first and second series,
respectively. The level of applied axial load (P/fcAg), the shear span-to-depth ratio of
each column (L/h), the configuration of transverse reinforcement, and anchorage details
of that reinforcement were the main variables. Table 1.3 illustrates the mechanical
properties of the materials and other details of the column specimens.
The plastic hinge region of the column specimens was designed according to the
NZS 3101:1982 code for both confinement and shear. After testing the two series of
column units, the equivalent plastic hinge lengths were found to be between 0.40 and
0.75 of the overall depth of the column section. Tanaka and Park (1990) observed that
when the axial load level increased, the equivalent plastic hinge length increased.
18
Table 1.3. Details of column specimens (Tanaka and Park 1990)

Longitudinal
Steel
fc,
fy,
sh,
ksi
l
ksi
in.
(Mpa) P/fcAg (MPa) (%) (mm)
(2)
Unit
(1)
(4)
(3)
(5)
3.71
68.73
3.15
1
0.2
1.57
(26)
(474)
(80)
3.71
68.73
3.15
2
0.2
1.57
(26)
(474)
(80)
3.71
68.73
3.15
3
0.2
1.57
(26)
(474)
(80)
3.71
68.73
3.15
4
0.2
1.57
(26)
(474)
(80)
4.64
74.1
4.33
5
0.1
1.25
(32)
(511)
(110)
4.64
74.1
4.33
6
0.1
1.25
(32)
(511)
(110)
4.65
74.1
3.54
7
0.3
1.25
(32)
(511)
(90)
4.65
74.1
3.54
8
0.3
1.25
(32)
(511)
(90)
Transverse Steel
fyh,
ksi
(MPa)
(6)
48.29
(333)
48.29
(333)
48.29
(333)
48.29
(333)
47.13
(325)
47.13
(325)
47.13
(325)
47.13
(325)
s
(%)
(7)
Ash/Ash,ACI
(8)
bxh,
in. (mm)
(9)
L/h
(10)
2.55
1.06
15.75x15.75
4.00
2.55
1.06
15.75x15.75
4.00
2.55
1.06
15.75x15.75
4.00
2.55
1.06
15.75x15.75
4.00
1.70
0.82
21.65x21.65
3.00
1.70
0.82
21.65x21.65
3.00
2.08
1.00
21.65x21.65
3.00
2.08
1.00
21.65x21.65
3.00
Lp,
in.
(mm)
(11)
6.77
(172)
8.7
(221)
10.6
(269)
11.06
(281)
8.35
(212)
13.66
(347)
14.45
(367)
18.66
(474)
2: Applied axial load ratio

5: Spacing of transverse reinforcement
9: Cross sectional dimensions
10: Shear-span-depth ratio
11: Measured plastic hinge length
1.3.8
Paulay and Priestley (1992)

Paulay and Priestley (1992) reported that theoretical values for the equivalent
plastic hinge length based on integration of the curvature distribution for typical
members should be dependent on l, where l is the column height. Plastic hinge length
values that are needed to calculate a measured lateral displacement, however, were not
19
consistent with experimentally measured lengths of lp. As Fig. 1.5 shows, the theoretical
curvature distribution ends abruptly at the base of the cantilever (Fig. 1.5b), whereas the
actual steel tensile strains should continue for some depth into the footing due to finite
bond stress. The elongation of longitudinal bars beyond the theoretical base causes
additional rotation and deflection (tensile strain penetration as shown in Fig. 1.5c). The
following formula was proposed by revising Eq. (1.16) to consider the effect of flexural
reinforcement with different strengths on the length of a plastic hinge formed at the
bottom of a cantilever column:
l p = 0.08l + 0.15d b l f ye 0.3d b l f ye
(ksi)
(1.17a)
(MPa)
(1.17b)
where l is the height of the cantilever column, fye is the yield stress of longitudinal
reinforcement, and dbl is the diameter of the longitudinal reinforcement.
Paulay and Priestley (1992) recommended that Eq. (1.17) results in values of lp
close to 0.5d, where d is the section depth, for typical beams and columns. It was
observed that the equivalent plastic hinge length and the region of plasticity where
special reinforcing detailing is required must be defined separately to ensure dependable
inelastic rotation capacity. This difference is shown in Fig. 1.5b by indicating the spread
of plasticity over a region outside the equivalent plastic hinge length.
20
(a) Yield curvatures
(b) Curvature at
(c) Equivalent curvatures
maximum response
Figure 1.5 Theoretical curvature relationships for a prismatic reinforced concrete
cantilever column (Paulay and Priestley 1992)
1.3.9
Soesianawati, Park and Priestley (1986), Watson and Park (1994)

Soesianawati et al. (1986) conducted experimental research on four square
concrete columns under low axial loads. The column specimens were designed with
smaller quantities of confining reinforcement than those recommended by the
NZS3101:1982 code (New Zealand Standards). Watson and Park (1994) furthered the
experimental research of Soesianawati et al. (1986) by testing five more square columns
and two octagonal columns under moderate to high axial compression load levels. Table
1.4 gives the details of the square column specimens, which have 16 in. (400 mm) square
cross sections and a height of 64 in. (1600 mm). The shear span-to-depth ratio was 4 for
the test specimens.
Units 1 to 4 were subjected to low axial load (P = 0. 1fcAg to 0. 3fcAg). Units 1
and 2 contained 43% and 46% of the New Zealand code recommended quantity of
transverse reinforcement. These specimens reached displacement ductility factors of at
21
least 8 without significant strength degradation, where displacement ductility factor is

the ratio of lateral displacement to the displacement at first yield. Unit 3, having 30% of
the code required quantity of transverse reinforcement, achieved a displacement ductility
factor of 6. Unit 4, designed with 17% of the code recommended quantity of transverse
reinforcement, was capable of reaching a displacement ductility factor of 4 and showed
hoop anchorage failure and buckling of longitudinal bars.
Units 5 and 6 were tested under high axial load with P = 0.5fcAg. These
specimens, with 38% and 19% of the confining reinforcement required by the New
Zealand code, achieved displacement ductility factors of 6.7 and 5.4, respectively. At the
end of the test, buckling of longitudinal bars was observed. The axial load level of P =
0.7fcAg was applied to Units 7, 8 and 9, which contained 48%, 34%, and 93% of the
code recommended quantity of confining reinforcement for ductile detailing,
respectively. Units 7 and 8 achieved displacement ductility factors of 6.3 and 4.0,
respectively. Unit 9 showed remarkably good performance and the test was continued
until reaching the displacement ductility factor of 10.
Watson and Park (1994) observed that the length of potential plastic hinge
regions increased as the axial load level increased. The other parameters, such as the
aspect ratio and the section type of the columns, were found not to have a
significant effect. The equivalent plastic-hinge length was calculated using the Eq.
(1.16) for column units and found to be 0.56*h, where h= column depth. The NZS 3101recommended that confined length was insufficient for many columns, particularly for
those with large axial compression.
They proposed the following formula (Eq. 1.18) to calculate the length of the
confined region for most columns:
22
lc
P
= 1 + 0.4 '
h
f c Ag
(ksi)
(1.18a)
lc
P
= 1 + 2.8 '
h
f c Ag
(MPa)
(1.18b)
where
lc = length of confined region of column, in. (mm)
h = lateral dimension of rectangular column section, in. (mm)
= strength reduction factor

fc = compressive cylinder strength of concrete, ksi (MPa)
Ag = gross area of column section, in.2, (mm2).
It is recommended that Eq. (1.18) be used in design. This expression gives lc equal to h
when the axial load is zero, and lc equal to 3h when the axial load is 0.70Agfc.
23
Table 1.4. Details of column specimens (Watson and Park 1994)

Axial Load
Longitudinal
Reinforcement
fc ,
P,
fy,
ksi
kips
ksi
(MPa) (kN) P/fcAg (MPa)
(1)
(2)
(4)
Unit
(3)
6.74
167
64.67
1
0.1
(46)
(743)
(446)
6.38
475
64.67
2
0.3
(44)
(2113)
(446)
6.38
475
64.67
3
0.3
(44)
(2113)
(446)
5.8
432
64.67
4
0.3
(40)
(1922)
(446)
5.95
737
68.73
5
0.5
(41)
(3278)
(474)
5.8
719
68.73
6
0.5
(40)
(3198)
(474)
6.09
1058
68.73
7
0.7
(42)
(4706)
(474)
5.65
982
68.73
8
0.7
(39)
(4368)
(474)
5.8
1007
68.73
9
0.7
(40)
(4479)
(474)
l
(%)
(5)
1.51
1.51
1.51
1.51
1.51
1.51
1.51
1.51
1.51
Transverse Reinforcement
d b,
in.
(mm)
(6)
0.28
(7)
0.32
(8)
0.28
(7)
0.24
(6)
0.32
(8)
0.24
(6)
0.47
(12)
0.32
(8)
0.47
(12)
fyh,
ksi
(MPa)
(7)
82
(565)
81
(558)
82
(565)
57
(393)
84
(579)
87
(600)
69
(476)
84
(579)
69
(476)
s
(%)
(8)
Ash/Ash,ACI
(9)
lp/h
(Eq.
1.13)
(10)
0.84
0.36
0.56
1.2
0.55
0.56
0.79
0.36
0.56
0.56
0.2
0.56
1.15
0.58
0.56
0.55
0.29
0.56
2.16
0.9
0.56
1.21
0.64
0.56
3.99
1.75
0.56

6: The diameter of longitudinal reinforcement
10: The ratio of calculated plastic hinge length using Eq. 1.16 to the depth of the column
1.3.10 Sheikh and Khoury (1993), Sheikh, Shah and Khoury (1994)
Sheikh and Khoury (1993) and Sheikh et al. (1994) completed experimental
research on six large-scale normal-strength concrete and four high-strength concrete
column specimens. The concrete columns were 72.5 in. (1842 mm) high and had 12-in.
(305 mm) square cross sections that result in a shear span-to-depth ratio of 6. The
24
concrete strength, level of axial load, and the percentage of transverse reinforcement
were the main test variables. Table 1.5 shows the details of the specimens tested and the
applied axial load.
The primary goal of the research was to assess the confinement provisions of the
ACI 318 (1989) code. According to this version of the code, the total cross sectional area
of rectangular hoop reinforcement for confinement (Ash) should not be less than that
given by the following Eq. (1.19a and b):
Ag
f c'
Ash = 0.3sbc
1
A
ch
f yt
Ash = 0.09 sbc
f c'
f yt
(1.19a)
(1.19b)
where
Ag = gross area of column section, in.2 (mm2)
Ach = area of core concrete measured out-to-out of transverse reinforcement, in.2
(mm2)
fc = compressive strength of concrete, ksi (MPa)
fyt = yield strength of transverse reinforcement, ksi (MPa)
s = spacing of transverse reinforcement, in. (mm)
bc = cross sectional dimension of column core, measured center-to-center of
transverse reinforcement, in. (mm).
In the ACI 318 code, the length of the column requiring confinement is specified
as the greatest of the overall depth (h) of a column at the joint face (where h is the larger
sectional dimension for a rectangular column or the diameter of a circular column), onesixth of the clear height of a column, or 18 in. (457 mm). The spacing of transverse
25
reinforcement is required to be less than h/4 or 6db, where h is the minimum member
dimension and db is the diameter of longitudinal reinforcement.
The researchers concluded that a column designed according to the ACI (1989)
code requirements has adequate performance in terms of curvature and displacement
ductility, but only for certain situations. Depending on the reinforcement detailing and
axial load level, the code provisions may give unnecessarily conservative design. It was
also observed that the measured plastic hinge lengths were an average value of 1.0h in
the column tests as shown in Table 1.5, where h is the column depth. Most of the column
tests were, however, conducted under high axial loads. It also appeared that steel
configuration, axial load level, amount of confining steel, and concrete strength did not
have an influence on the plastic hinge length.
26
Table 1.5 Details of specimens (Sheikh and Khoury 1993, 1994)

Longitudinal Steel
fc,
ksi
(MPa)
(1)
No.
of
bars
(2)
FS-9
4.7
(32)
8
8
ES-13
4.72
(33)
AS-3
4.81
(33)
AS-17
4.54
(31)
4.75
(33)
Spec.
AS_18
4.68
(32)
l
(%)
(3)
fyl,
ksi
(MPa)
(4)
s
(%)
(5)
fyh,
ksi
(MPa)
(6)
2.44
73.6
(507)
1.68
2.44
73.6
(507)
P/fcAg
(8)
73.6
(507)
1.46
0.76
13.1
(333)
1.10
1.69
67.3
(464)
1.34
0.76
10.2
(259)
0.85
2.44
73.6
(507)
1.68
73.6
(507)
1.43
0.60
11.5
(292)
0.96
2.44
73.6
(507)
1.68
73.6
(507)
1.52
0.77
12.6
(320)
1.05
73.6
(507)
3.06
67.3
(464)
2.41
0.77
11.9
(302)
0.99
73.6
(507)
1.30
73.6
(507)
67
(462)
1.12
0.47
13.9
(353)
1.16
2.44
73.6
(507)
1.68
73.6
(507)
0.88
0.62
12.7
(323)
1.05
2.44
73.6
(507)
3.06
67.3
(464)
1.44
0.64
10.7
(272)
0.89
2.44
73.6
(507)
4.30
67.3
(464)
2.10
0.64
13
(330)
1.08
2.44
73.6
(507)
1.68
73.6
(507)
0.80
0.65
2.44
8
AS-19
2.44
AS3H
7.86
(54)
AS18H
7.93
(55)
AS20H
7.78
(54)
A-17H
8.57
(59)
Lp,
in.
(mm)
(9)
Ash/
Ash,ACI
(7)
Lp/h
(10)

2: Number of bars used in the specimens
7: The total cross sectional area of rectangular hoop reinforcement for confinement according to ACI 318 (1989)
9: The measured plastic hinge length
10: The ratio of measured plastic hinge length to the column depth
*: No. 3 (10) and 6-mm bars were used for the perimeter ties and inner ties, respectively
1.3.11 Kovacic (1995)

As part of a long-term study on the behavior of high-strength concrete structures
at the University of Melbourne, Kovacic (1995) conducted an experimental and
27
theoretical investigation of the full-range behavior of high-strength concrete columns,

with the nominal concrete strength as high as 11.6 ksi (80 MPa), and with low axial load
ratios ranging between 5% and 20%. Six out of eight column test results (Table 1.3) were
within the ACI limits for plastic hinge length given in 1968 (Eq. 1.7 and 1.8), and thus
justified using these equations to estimate the hinge lengths for high-strength concrete
columns with low axial loads. Kovacic reported that the ACI formulae gave reliable
predictions of hinge lengths for high strength concrete columns with low axial loads, but
more experiment was required to confirm and extend these observations for columns
with high axial loads and for very high concrete strengths.
Table 1.6. Details of beams tested by Kovacic
Concrete
Span,
Width,
Depth,
Axial
strength
in.
in.
in.
force,
fc, ksi
Measured
ACI 1ACI 2(mm)
(mm)
(mm)
kips (kN)
Label
(MPa)
Lp/d
Lower/d
Upper/d
50
3.15
5.91
14.61
4.82
0.254
(1270)
(80)
(150)
(65)
(33)
D1
0.37
0.91
50
3.15
5.91
14.61
4.74
0.467
D2
(1270)
(80)
(150)
(65)
(33)
0.37
0.91
50
3.15
5.91
14.61
5.61
0.633
D3
(1270)
(80)
(150)
(65)
(39)
0.37
0.91
50
3.15
5.91
29.22
8.43
0.299
D4
(1270)
(80)
(150)
(130)
(58)
0.37
0.91
50
3.15
5.91
14.61
9.33
0.467
D5
(1270)
(80)
(150)
(65)
(64)
0.37
0.91
50
3.15
5.91
29.22
9.4
0.699
D6
(1270)
(80)
(150)
(130)
(65)
0.37
0.91
50
3.15
5.91
7.19
9.11
0.547
(1270)
(80)
(150)
(32)
(63)
D7
0.37
0.91
50
3.15
5.91
14.61
9.46
0.467
D8
(1270)
(80)
(150)
(65)
(65)
0.37
0.91
1: The ratio of plastic hinge length calculated using ACI lower limit (1968) to the depth of the column
2: The ratio of plastic hinge length calculated using ACI upper limit (1968) to the depth of the column
1.3.12 Bayrak and Sheikh (1997, 1999)

Bayrak and Sheikh (1997) and Bayrak (1999) constructed and tested twenty four
square and rectangular concrete column specimens to study the effect of high-strength
28
concrete columns on plastic hinge length. The concrete strength for standard cylinders
ranged between 10,000 and 16,000 psi (72 MPa and 112 MPa). The cross-sections of the
columns were 12 in. (305 mm) square, and 12 in. (305 mm) by 10 in. (250 mm)
rectangular dimensions with 72.5 in. (1,841 mm) in height. The shear span-to-depth
ratios were 6, 7.4 and 5.3.
The plastic hinge lengths of the specimens tested were calculated using the Eq.
(1.2) for all the load cycles in which the displacement ductility factor is greater than 4
and then averaged to find the equivalent plastic hinge length for the columns. The
experimental plastic hinge lengths were close to the depth of column sections (h) as
given in column (9) of Table 1.7. It was suggested that a simpler expression such as Lp =
x*h, where x can have a value between 0.9 and 1, is more appropriate to obtain the
plastic hinge length for the columns.
In the two studies, as the axial load increased, the deformability of the reinforced
concrete columns reduced and strength and stiffness degradation with every load cycle
accelerated. Thus, a larger amount of lateral reinforcement was needed to balance this
effect. Bayrak and Sheikh (1997) concluded that the axial load level should be
considered in the design of confining reinforcement. Based on the test results, the
displacement ductility factors decreased with increasing shear span-to-depth ratios (L/h).
It was observed that section geometry and shear span-to-depth ratio influenced the
member-level ductility parameters (which are the displacement ductility factor, and work
damage indicator that was represented by the work done on the column by lateral load),
whereas section-level ductility parameters (which are the curvature ductility factor, and
energy damage indicator that was defined by energy dissipated in the plastic hinge
region) were not affected by these factors.
29
Table 1.7. Details and test results of column specimens
Unit
ES-1HT
AS-2HT
AS-3HT
AS-4HT
AS-5HT
AS-6HT
AS-7HT
ES-8HT
RS-9HT
RS10HT
RS11HT
RS12HT
RS13HT
RS14HT
RS15HT
RS16HT
RS17HT
RS18HT
RS19HT
RS20HT
WRS21HT
WRS22HT
WRS23HT
WRS24HT
fc, ksi
(MPa)
(1)
10.45
(72)
10.4
(72)
10.41
(72)
10.43
(72)
14.76
(102)
14.78
(102)
14.79
(102)
14.82
(102)
10.32
(71)
10.31
(71)
10.27
(71)
10.28
(71)
16.25
(112)
16.25
(112)
8.15
(56)
8.15
(56)
10.74
(74)
10.74
(74)
10.76
(74)
10.76
(74)
13.24
(91)
13.24
(91)
10.47
(72)
10.47
(72)
P/fcAg
(2)
0.5
0.36
0.5
0.5
0.45
0.46
0.45
0.47
0.34
0.5
0.51
0.34
0.35
0.46
0.36
0.37
0.34
0.5
0.53
0.34
0.47
0.31
0.33
0.5
Longitudinal
Steel
fy, ksi
l
(MPa) (%)
(3)
(4)
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
Transverse Steel
Spec.,in. fyh, ksi
s
(mm)
(MPa) (%)
(6)
(5)
(7)
3.74
67.14
(95)
(463) 3.15
3.54
78.59
(90)
(542) 2.84
3.54
78.59
(90)
(542) 2.84
3.94
67.14
(100)
(463) 5.12
3.54
78.59
(90)
(542) 4.83
2.99
67.14
(76)
(463) 6.72
3.7
78.59
(94)
(542) 2.72
2.76
67.14
(70)
(463) 4.29
3.15
78.59
(80)
(542) 3.44
3.15
78.59
(80)
(542) 3.44
3.15
78.59
(80)
(542) 5.43
5.91
78.59
(150)
(542) 1.83
2.76
67.43
(70)
(465) 3.92
2.76
67.43
(70)
(465) 3.92
3.94
67.43
(100)
(465) 2.75
5.91
67.43
(150)
(465) 1.83
2.95
197.2
(75)
(1360) 1.83
2.95
197.2
(75)
(1360) 1.83
2.95
203.29
(75)
(1402) 3.54
5.51
203.29
(140)
(1402) 1.9
2.76
67.43
(70)
(465) 3.92
2.76
67.43
(70)
(465) 3.92
3.15
78.59
(80)
(542) 3.44
3.15
78.59
(80)
(542) 3.44
30
Section Exp.
Depth,
Lp,
in.
in.
Ash/
(mm)
(mm)
Ash,ACI
(9)
(10)
(8)
12.01 13.82
1.13
(305)
(351)
12.01 11.73
1.19
(305)
(298)
12.01 10.87
1.19
(305)
(276)
12.01 10.71
(305)
(272)
1.83
12.01 10.31
1.08
(305)
(262)
12.01 12.64
1.62
(305)
(321)
12.01 10.55
0.8
(305)
(268)
12.01 15.16
1.08
(305)
(385)
13.78 14.09
(350)
(358)
1.72
13.78 17.48
1.72
(350)
(444)
13.78 15.59
2.29
(350)
(396)
13.78 16.42
0.92
(350)
(417)
13.78 11.65
1.09
(350)
(296)
13.78 13.82
(350)
(351)
1.09
13.78 10.71
1.49
(350)
(272)
13.78 14.84
1
(350)
(377)
13.78 11.65
1.39
(350)
(296)
13.78 12.56
1.39
(350)
(319)
13.78 13.54
2.67
(350)
(344)
13.78 13.74
(350)
(349)
1.43
9.84
11.02
1.31
(250)
(280)
9.84
10.98
1.31
(250)
(279)
9.84
10.08
1.7
(250)
(256)
9.84
9.72
1.7
(250)
(247)
Average
Standard Deviation
Lp/h
(11)
1.15
0.98
0.91
0.89
0.86
1.05
0.88
1.26
1.02
1.27
1.13
1.19
0.85
1
0.78
1.08
0.85
0.91
0.98
1
1.12
1.11
1.03
0.99
1.01
0.13
Notes to Table 1.7:

5: Spacing of transverse reinforcement
8: The total cross sectional area of rectangular hoop reinforcement for confinement according to ACI 318
9: Cross sectional depth
11: The ratio of measured plastic hinge length to the column depth
*: The maximum average tie strain reached in Specimens RS-17HT, RS-18HT, RS-19HT and RS-20HT is
0.00425. Therefore, maximum attainable strength of 850 MPa is used in the calculations.
1.3.13 Bae (2005)

Based on previous developed work [Bayrak and Sheikh (1997), and Bayrak
(1999)], a new experimental program was designed by Bae at the University of Texas at
Austin to investigate the influence of certain parameters on plastic hinge length. These
parameters were shear span-to-depth ratio (L/h), axial load level (P/Po), and amount of
confining reinforcement (Ash).
In this experimental program, four of the test specimens had column crosssection dimensions of 24 in. (610 mm) by 24 in. (610 mm) and a height of 103.5 in.
(2,630 mm), with end stubs having cross-section dimensions of 38 in. (965 mm) by 38
in. (965 mm) and a height of 80 in. (2030 mm). The dimensions of only one specimen
called S17-3UT had a 17.25 in. (440 mm) square cross section. Figure 1.6 illustrates a
typical specimen. Table 1.8 lists the details of the specimens and the axial load level. The
specified nominal 28 day strength of concrete was 4,000 psi (28 MPa) for the first
specimen, which was used as a guide to check the performance of the test setup, and
6,000 psi (42 MPa) for the other test specimens.
31
Table 1.8. Details of test specimens

Longitudinal Steel
bxh, in.
fc,
x in.
ksi
(mm x
mm)
(MPa)
(2)
Specimen
(1)
24 x 24
4.3
(610 x
(30)
S24-1UT
610)
24 x 24
6.3
(610 x
(43)
610)
S24-2UT
17.25 x
6.3
17.25
(43)
(438 x
S17-3UT
438)
24 x 24
5.3
(610 x
(37)
610)
S24-4UT
24 x 24
(610 x
6 (41)
S24-5UT
610)
1: Cross sectional dimensions
Bar
Size,
(SI)
(3)
No.9
(29)
No.71
(22)
l
(%)
(4)
2.08
1.25
No.5
(16)
fyl,
ksi
(MPa)
(5)
84
(579)
73
(503)
s
(%)
(6)
fyh,
ksi
(MPa)
(7)
1.28
2.04
72
(496)
1.25
No.72
(22)
No.72
(22)
Transverse Steel
1.25
1.25
64
(441)
62
(427)
58
(400)
P/Po
(9)
1.04
0.5
1.09
0.5
1.12
0.5
0.44
0.2
0.74
0.2
72
(496)
1.76
58
(400)
Ash/
Ash,ACI
(8)
0.72
1.3
66
(455)
63
(434)

3: Bar sizes for the longitudinal reinforcement according to English and SI units
8: The total cross sectional area of rectangular hoop reinforcement for confinement according to ACI 318
9: The ratio of axial load, where P0 = 0.85 fcAc + fylAs
32
Figure 1.6 Layout of Test Specimens (Bae 2005)
The behaviors that were observed from specimens S24-2UT and S17-3UT were
compared in detail to examine the effect of shear span-to-depth ratio (L/h) on the
response of concrete columns. Specimens S24-2UT and S17-3UT had 24 in. (610 mm)
and 17.25 in. (440 mm) square cross-sections, respectively. These two specimens were
designed with similar concrete strength, amount and detail of longitudinal and transverse
reinforcement. The only difference between them was the shear span-to-depth ratio for
specimens S24-2UT and S17-3UT due to different section sizes. The shear span-to-depth
ratios of these specimens were calculated as 5 and 7, respectively. Similar sectional
performance was obtained from tests of S24-2UT and S17-3UT. The curvature ductility
factors, which are the ratio of ultimate curvature to the curvature at first yield, were
33
calculated from the backbone curves of specimens S24-2UT and S17-3UT as 8.8 and
9.1, respectively. The member-level deformation capacities of the specimens, however,
were quite different. It was observed that as the shear span-to-depth ratio increased from
5 to 7, displacement ductility was significantly decreased from 5.9 to 2.7. The drift
capacity then dropped from 2.6% to 1.6%. From all the observations above, it can be
summarized that the shear span-to-depth ratio had an important effect on the overall
member performance of the test columns.
Bae (2005) introduced an analytical approach to estimate the plastic hinge length
of a concrete column. In this approach, first the intensity of concrete compressive strains
is measured at critical sections along the length of a column, and then using the
compressive strain profile, the yielding region length of longitudinal bars can be
estimated. Finally, the plastic hinge length is calculated by subtracting the stub
confinement effect (approximately 0.25h where h is the column depth) from the overall
length where the compressive bar strains are greater than the yield strain.
A series of theoretical analyses were carried out to study the influence of the
shear span-to-depth ratio on the plastic hinge length. A 24 in. (610 mm) square column
having different shear span-to-depth ratios (L/h) was designed. The longitudinal
reinforcement was constant as l = 0.01. The results of the analyses are shown in Fig.
1.7. The plastic hinge length increased with increasing L/h, and this effect was more
pronounced as the axial load increased.
34
Figure 1.7 Relationship between Plastic Hinge Length and Shear Span-to-Depth Ratio
(Bae 2005)
To verify the conclusions observed in the analytical study, specimens S24-4UT

and S24-5UT [which were tested under low axial loads (P = 0.2Po)] were examined in
greater detail to study the effect of axial load. The behaviors of the specimens S24-4UT
and S24-5UT were compared with the behavior of specimen S24-2UT that was tested
under a high axial load level of 0.5Po. The confinement reinforcement of specimens S244UT and S24-5UT were not designed according to the ACI 318-05 code requirements,
whereas specimen S24-2UT met the code requirements. The shear span-to-depth ratio
(L/h=5) was the same for these specimens. It was observed that the deformation capacity
of specimen S24-2UT was less than that of specimens S24-4UT and S24-5UT. The
lateral load-displacement responses were normalized to maximum lateral load and yield
displacement. The normalized response of S24-4UT, however, showed less ductile
response than that of specimen S24-2UT. The member-level ductility parameters
35
(displacement ductility, and energy absorption) for specimen S24-2UT gave higher
values than those of specimen S24-4UT, but less than those of specimen S24-5UT.
To determine the sensitivity of the analysis, a 24 in (610 mm) square column with
120 in. (3,050 mm) length was studied to investigate the effect of axial load on the
plastic hinge length. The percentage of longitudinal reinforcement was 1% for each
specimen. Concrete strength was 6,000 psi (42 MPa). The yield and ultimate strength of
reinforcing bars were 60 ksi (414 MPa) and 90 ksi (620 MPa), respectively. Along the
length of columns, the curvature and compressive concrete strain profiles were examined
for different axial load levels (from 0.2Po to 0.5Po).
Figure 1.8 shows the effect of axial load on curvature and compressive strain of
concrete. The curvature profiles are not affected from the various axial load levels
whereas the compressive strain profiles indicate the influence of axial load very well.
Figure 1.8 Effect of Axial Load on Curvature and Compressive Strain Profiles (Bae
2005)
To estimate the plastic hinge length, the region of reinforcing bars that yield in
compression was investigated along the column. Fig. 1.9 shows that the plastic hinge
36
length is nearly constant for low axial loads (P 0.2Po) where it is approximately equal
to h/4. The plastic hinge length increases as axial loads increase (P>0.2Po).
Figure 1.9 Relationship between Plastic Hinge Length and Axial Load (Bae 2005)
The estimated plastic hinge length is approximately equal to 0.8h at the axial load
level of 0.52Po, which is the maximum axial load allowed by ACI 318-05 (Pn,max =
0.8Po = 0.650.8Po=0.52Po for tie reinforcement). An additional length of 0.25h
should be added to the plastic hinge to take into account the stub confinement effect. The
required confined length was calculated as 1.05h by adding the largest estimated plastic
hinge length and a distance of 0.25h. The length of the region of confinement should be
the largest of the overall depth of a column, one-sixth of the clear height of a column, or
18 in. (457 mm) according to Chapter 21 of the ACI 318-05 code. The confinement
region length indicated by ACI 318-05 appears to be satisfactory for low axial loads.
To evaluate the effect of longitudinal reinforcement on the plastic hinge lengths,
a 24 in. (610 mm) square column with 120 in. (3,050 mm) length was designed. The
shear span-to-depth ratio of the column specimens was 5. The curvature and compressive
37
strain profiles along the length of the column were examined to obtain the plastic hinge
lengths. Results showed that as the longitudinal reinforcement ratio (l) increases, the
length of the plastic hinge increases. Figure 1.10 shows the results for different axial load
levels.
Figure 1.10 Effect of Amount of Longitudinal Reinforcement (Bae 2005)
After all the sensitivity analyses were done, results showed that axial load, shear
span-to-depth ratio, and the amount of longitudinal reinforcement significantly
influenced the length of plastic hinges. Using linear relationships between the parameters
(P/Po, L/h, and As/Ag) and the plastic hinge, a plastic hinge length expression was
calibrated based on the analysis results. Equation 1.25 was proposed from the result of a
series of least squares analyses that were done on the UW/PEER column database
(http://maximus.ce.washington.edu/~peera1/).
P
= 0.3
h P0
lp
As
+ 3
Ag
0.1 L + 0.25 0.25
38
(1.20)
The plastic hinge lengths obtained from the analysis of compressive strains along
the column height and those from using Eq. 1.20 were compared in Fig. 1.11. It was
observed that the plastic hinge lengths estimated by the analytical approach and Eq. 1.20
did not match very well, especially at high axial loads and large shear span-to-depth
ratios, because the plastic hinge lengths obtained from the analysis of compressive
strains are approximate results. It was important to mention that Bae (2005) proposed to
find a simple expression so that conservative estimations for the deformation capacity of
concrete columns can be obtained. It can be summarized that moderate plastic hinge
length values can be determined using the new formula for a wide range of axial loads,
shear span-to-depth ratios, and amount of longitudinal reinforcement.
(a) Axial Load
(b) Shear Span-to-Depth Ratio
(c) Amount of Longitudinal Reinforcement

Figure 1.11 Comparisons of Plastic Hinge Length (Eq. (1.20) versus Analysis) (Bae
2005)
39
Table 1.9. Comparisons of measured and proposed plastic hinge lengths

Specimen
Measured
Lp
Proposed
Lp
S24-1UT
S24-2UT
S17-3UT
S24-4UT
S24-5UT
0.66h
0.91h
0.49h
0.47h
0.69h
0.86h
0.25h
0.25h
1.3.14 Restrepo, Seible, Stephan, and Schoettler (2006)

To assess the seismic performance of columns using high-strength reinforcing
steel, two 35%-scale circular bridge column units were constructed and tested under
quasi-static reversed cyclic loading at University of California, San Diego. Unit 1 was
built with Grade 60 ASTM A 706 reinforcement whereas Unit 2 was built with highstrength reinforcing steel (94 ksi or 650 MPa) to compare the behavior of columns.
The ultimate displacement capacity for the two tested units was estimated using
the equivalent plastic hinge length given by Eq. (1.17) (Paulay and Priestley 1992) and
found to be less than the values measured during the tests. The equivalent plastic hinge
length given by Eq. (1.17) was subsequently modified using two generic multipliers
and
L p = H + f ye d bl 2 f ye d bl
(1.21)
where is a yield spread coefficient and is a strain penetration coefficient.

For a cantilever column (Fig. 1.12(a)), the flexural component of lateral
deflection at the top of the cantilever fle was determined from the integration of the
curvature distribution in the column. Fig.1.12(c) shows the moment and curvature
distributions along the column height.
H
fle = ydy = elf + pf

0
40
(1.22)
The resulting deflection can be divided into elastic elf (Eq. 1.23a) and plastic pf
(Eq.1.23b) components that are consistent with the equivalent curvature distribution
drawn in Fig. 1.12(d). This calculation is demonstrated in Eq. (1.23):
M ' H2
y
M
y
elf =
(1.23a)
pf = pH 2 (1 )
(1.23b)
where p is the plastic curvature given by Eq. (1.24):
p =
M '
y
My
(1.24)
Combining Eq. (1.22) through (1.24), and finding the solution for
fle
= 1 1 2
(a)
M ' H2
(
y )
My
3
(1.25)
pH 2
(b)
(c)
(d)
Figure 1.12 Idealization of curvature distribution in column: (a) column; (b) BMD; (c)
curvature diagram; and (d) equivalent curvature diagram (Restrepo et al. 2006)
The strain penetration coefficient can be obtained from the reading of the first
and second displacement transducers that were instrumented immediately above the
41
column base. The fixed-end rotation shown in Fig. 1.12(a) can be calculated from the
first set of displacement transducers if the gauge length is small enough to minimize the
effect of local bond slip in the rotation. The second set of displacement transducers is
used to estimate the base curvature shown in Fig. 1.12(d). The plastic hinge length due
to strain penetration can be written as
f ye d bl =
(1.26)
Therefore, is
(1.27)
f ye d bl
Figure 1.13 and 1.14 illustrate and versus the curvature ductility demand
gained from both test units, respectively. The values of = 0.08 and = 0.15 (if fye is in
MPa, is equal to 0.022) used in Eq. (1.17) are shown in these figures.
It can be seen that the coefficient increases as the curvature ductility demand
increases because the spread of plasticity takes place gradually until a column reaches its
maximum ductility capacity. Figure 1.13 shows that the maximum spread of plasticity is
significantly greater than predicted by Paulay and Priestley (1992) for both test results.
Unlike the spread of plasticity coefficient , the strain penetration coefficient is
absolutely independent from the curvature ductility demand as shown in Fig. 1.14. It is
seen that the values of coefficient obtained for Units 1 and 2 are also greater than the
value proposed by Paulay and Priestley (1992). It is expected that equivalent plastic
hinge lengths will be greater than those obtained by Eq. (1.17) because both values of
coefficients are greater than the values recommended by Paulay and Priestley (1992).
Therefore, the theoretical lateral deformation capacity is greatly underestimated for
circular columns when using Eq. (1.17).
42
Restrepo et al. (2006) observed that using higher-strength reinforcement could be

promoted in seismic design. In addition, using higher-strength transverse reinforcement
restrains the longitudinal reinforcement and prevents buckling. Restrepo et al. observed
that the equivalent plastic hinge lengths determined from the test units were greater than
the equivalent length predicted by Eq. (1.17) (Paulay and Priestley 1992) by at least 40%
as shown in Fig. 1.15.
Figure 1.13 Plasticity spread coefficient ((Restrepo et al. 2006)
43
Figure 1.14 Strain penetration coefficient (Restrepo et al. 2006)
Figure 1.15 Equivalent plastic hinge length as ratio of column diameter (Restrepo et al.
2006)
44
1.3.15 Phan V., Saiidi M.S., Anderson J., and Ghasemi H. (2007)
Phan et al. (2007) designed two one-third scale reinforced concrete bridge
columns (labeled NF-1 and NF-2) that were tested on a shake table at the University of
Nevada, Reno Large Scale Structures Laboratory to investigate near-fault ground motion
effects on bridge columns. NF1 and NF2 had the same cross-sectional properties and
column height. The only difference was that NF1 was designed according to 2004
Caltrans Seismic Design Criteria (SDC) version 1.3, whereas NF2 was designed based
on the AASHTO 2002 Standard Specifications for Highway Bridges (AASHTO 2002).
NF1 column specimen was nearly identical to 9F1 that was tested in the same laboratory
during a previous study (Laplace 2005). The only difference between NF-1 and 9F1 was
that NF-1 was subjected to near-fault impulsive ground motion whereas 9F1 was tested
under an earthquake record that did not include forward directivity effects (El Centro
1940).
Paulay and Priestleys equation (Eq. 1.17) was used to estimate the theoretical
plastic hinge length (lp) for each specimen. The theoretical values are generally
conservative for conventional reinforced concrete columns. For all three specimens, the
theoretical lp was calculated as 11.26 in. (286 mm). Table 1.10 lists the measured lp for
all the test specimens to examine whether current methods for calculating lp may
underestimate the actual value in structures subjected to near-fault ground motions
(Hamilton et al. 2001). The measured lp were 51-95% longer than those determined using
Eq. (1.17). The lp for specimen NF-2 was very similar to that of specimen 9F1, and lp for
NF-1 was also comparable. Phan et al. concluded that there did not appear to be enough
evidence to make alterations of lp based on these tests.
45
Table 1.10 Measured plastic hinge lengths
Span/depth
L/d
Plastic
hinge
length
lp, in.
(mm)
Ratio to
diameter lp/d
68.01
(469)
4.5
22 (559)
1.38
6.15
(42)
68.01
(469)
4.5
17.05
(433)
1.07
6 (41)
68.01
(469)
4.5
17.01
(432)
1.06
Specimen
Column
diameter
d, in.
(mm)
Concrete
strength
f'c, ksi
(MPa)
Steel
strength
fy, ksi
(MPa)
NF-1
16 (406)
6 (41)
NF-2
16 (406)
9F1
16 (406)
1.3.16 Berry, Lehman, and Lowes (2008)

Berry et al. (2008) used the data from the tests of large-scale circular bridge
columns to evaluate the models for performance-based earthquake engineering
requirements for bridge columns, including a new expression for plastic hinge length.
Thirty-seven spiral-reinforced column tests out of 168 possible tests in the UW-PEER
Structural Performance Database (Berry 2004) were constructed with modern design
details after 1980. Each of these columns had the following design details, which were
used to define modern designed bridge columns:
Axial load ratio: P/fcAg 0.3, where P is the column axial load, f c' is the
concrete compressive strength, and Ag is the gross section area of the
column;
Spiral pitch 6db, where db is the diameter of the longitudinal

reinforcement;
Effective confinement ratio eff 0.05, where
eff =
s f ys
(1.28)
f c'
46
where s is the spiral reinforcement ratio which is the ratio of spiral

reinforcement volume to the volume of core concrete, and fys is the yield
strength of the spiral;
Depth of cover concrete over transverse reinforcement 0.1D, where D is

the diameter of the column; and
Longitudinal reinforcement ratio 4%.
Berry et al. (2008) evaluated the plastic hinge length expression provided by
Paulay and Priestley (1992) which has been adopted by California Department of
Transportation. A new expression was proposed by the authors:
l p = 1 D + 2 L + 3
f ydb
(1.29)
f c'
The last term in Eq. 1.29 was added to include the effect of strain penetration on the
spread of plasticity. The new expression was calibrated using the test data and the
parameters (1, 2, and 3) were determined using available optimization algorithms.
Two plastic hinge length expressions were defined as an optimal and a recommended
expression. The optimal expression was determined from the minimum value of the
errors, which were the errors associated with simulation of the column forcedisplacement response history, and the errors due to the ratio of measured-to-calculated
displacements at the beginning of spalling and bar buckling. The optimal expression is
written as:
l p = 0.0375 L + 0.01
f ydb
l p = 0.0375 L + 0.12
f y db
f c'
f c'
47
(psi)
(1.30a)
(MPa)
(1.30b)
A simplified recommended expression that represents a slight increase in error

(from 8.05 to 8.09) is given for design purposes:
l p = 0.05 L + 0.008
l p = 0.05 L + 0.1
f y db
f c'
f ydb
(psi)
(MPa)
f c'
(1.31a)
(1.31b)
The force-displacement response predicted using Eq. (1.31a and b) for columns
with different axial load and aspect ratios shows good comparison with the envelope to
the measured cyclic response, as shown for selected tests in Fig. 1.16.
(a) Columns with different axial load ratios
48
(b) Columns with different aspect ratios

Fig. 1.16 Simulated and observed force-drift ratio for columns with different design
parameters (Berry et al. 2008)
1.4
LITERATURE REVIEW OF PLASTIC HINGE LENGTH (DYNAMIC

TESTS)
1.4.1
Dodd et al. (2000)

Fourteen one-sixth scale single-stem reinforced concrete cantilever columns with
various aspect ratios and axial-load ratios were tested on a shaking table to investigate
the applicability of current design provisions and theoretical models that are based on
test data from quasi-statically tested specimens, to more realistic dynamic base excitation
conditions that columns experience during earthquakes.
The main parameters of interest were the aspect ratio of the columns, the axial
load ratio, the base flexibility, the base input motions, and the effect of initial low-level
shaking on the response of columns to subsequent higher-level shakes. Table 1.11 lists
details of all 14 columns. All of the columns were 200 mm in diameter and contained 18
49
uniformly distributed longitudinal bars having a 0.24 in. (6 mm) diameter with an
average yield stress of 65 ksi (450 MPa). The longitudinal steel ratio was 1.62%. The
compressive strengths at the time of testing were given in Table 1.11. Columns with
aspect ratios of 4, 7, and 10 were tested to provide data on a range of flexurally
dominated columns. Axial-load ratios of 0.05 and 0.4 were chosen to check the behavior
of columns above and below the balance point on the axial load moment interaction
diagram.
A record based on the 1940 El Centro N-S component earthquake was used as a
base motion for all columns except 1a, 2a, and 3a, which were subjected to a uniform
sinusoidal base motion.
The effective plastic hinge length on a bridge column was considered to be the
equivalent length over which all plasticity takes place with a constant plastic curvature.
The plastic hinge length was calculated using the equation proposed by Priestley et al.
(1992) and the reduction factor defined by Zahn et al. (1986). The reduction factor was
to be 0.5+1.67P/(fcAg) for P/(fcAg) less than 0.3 to modify Priestleys equation. It was
observed from the experimental values that axial load has an effect on the plastic hinge
length and the equation overestimates the plastic hinge length for the columns. It was
also determined from the comparisons of calculated and measured plastic hinge length
values that Zahns recommendation is mostly conservative and agrees well with the
measured values for all the columns having different aspect ratios.
50
Table 1.11 Details of the column properties

Longitudinal
Steel
fc, ksi
(MPa)
(1)
5.9
(41)
4.64
(32)
fy, ksi
l
(MPa) (%)
(4)
(5)
65
1a
0.04
1.62
(448)
65
1b
0.051
1.62
(448)
65
2a
5 (34) 0.048
1.62
(448)
4.39
65
2b
0.054
1.62
(30)
(448)
5.42
65
3a
0.044
1.62
(37)
(448)
4.22
65
3b
0.056
1.62
(29)
(448)
5.71
65
4
0.391
1.62
(39)
(448)
5.39
65
5
0.388
1.62
(37)
(448)
5.6
65
1.62
(39)
6
0.387 (448)
5.99
65
7
1.62
(41)
(448)
0.04
5.81
65
8
1.62
(40)
0.392 (448)
5.86
65
9
1.62
(40)
(448)
0.041
5.35
65
10
1.62
(37)
0.044 (448)
6.24
65
11
1.62
(43)
0.358 (448)
Column
No.
P/fcAg
(3)
Transverse
Steel
fyh, ksi
(MPa)
(6)
37.7
(260)
37.7
(260)
37.7
(260)
37.7
(260)
36.25
(250)
36.25
(250)
37.7
(260)
37.7
(260)
36.25
(250)
36.25
(250)
37.7
(260)
36.25
(250)
37.7
(260)
37.7
(260)
s
(%)
(7)
0.88
0.88
1.06
1.06
0.66
0.66
0.94
1.02
0.79
0.46
0.94
0.46
0.88
0.94
D,
in.
(mm)
(8)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)

8: Diameter of the columns
9: Aspect ratio
10: Calculated plastic hinge length using Eq. 1.17
51
L/D
(9)
4
4
7
7
10
10
4
7
10
4
4
7
4
4
Calc.
Lp,
in.
(mm)
(10)
4.84
(123)
4.84
(123)
6.73
(171)
6.73
(171)
8.62
(219)
8.62
(219)
4.84
(123)
6.73
(171)
8.62
(219)
4.84
(123)
4.84
(123)
6.73
(171)
4.84
(123)
4.84
(123)
Meas.Lp,
in. (mm)
(11)
3.54
(90)
3.35
(85)
4.92
(125)
5.31
(135)
7.68
(195)
5.91
(150)
4.72
(120)
11.02
(280)
10.83
(275)
3.35
(85)
7.48
(190)
5.31
(135)
3.74
(95)
7.68
(195)
1.4.2
Hachem et al. (2003)

The dynamic testing of four circular reinforced concrete bridge columns on the
earthquake simulator of the Earthquake Engineering Research Center at the University of

California Berkeley was conducted to study the behavior of ductile bridge columns
subjected to cyclic loading.
The specimens had a circular section with a diameter of 16 in. (405 mm) and the
aspect ratio was 6. The axial load in the columns was taken to be 0.1Agfc assuming a
typical Caltrans concrete design strength of 3.25 ksi (22.4 MPa). Since the actual
concrete compressive strength was 5.7 ksi (39.3 MPa) on average, the real axial load
ratio was about 0.06fcAg. The yield strength of longitudinal reinforcement was 60 ksi
(414 MPa). The longitudinal reinforcement consisted of 12 No. 4 deformed grade 60
(A706) reinforcing bars. The longitudinal and volumetric spiral reinforcement ratios
were 1.2% and 0.54%, respectively.
The specimens were divided into two groups. In each group, one specimen was
tested under one component of the ground motion, while the other specimen was
subjected to two components of the motion. The first pair was subjected to a modified
version of the Olive View record of the 1994 Northridge earthquake, which was a nearfault ground motion containing a high velocity pulse. The second pair was tested under
the 1985 Chile earthquake recorded at the Llolleo station, which was 40 miles (64 km)
away from the epicenter of the event.
The plastic hinge length before testing the circular bridge columns under
bidirectional earthquake loading was estimated using the equation proposed by Priestley
et al. (1992). The equation resulted in a plastic hinge length of 13.3 in. (338 mm). After
testing the columns, it was also observed that plastic hinge length ranged from 12 in.
52
(305 mm) to 14 in. (356 mm) which was about 0.75D, where D is the diameter of the
columns. The estimated plastic hinge length was within the observed range.
1.5
SUMMARY
Reinforced concrete bridge systems are generally designed to form plastic hinges
in columns rather than the beams in the superstructure. Special detailing must be
provided in reinforced concrete bridge columns along and beyond concentrated yielded
regions (designated as the plastic hinge zones) to provide satisfactory behavior during
response to strong ground motions. In addition, simple models to estimate drift in
reinforced concrete bridge columns can be developed using a definition of lumped
plasticity in the defined plastic hinge zone. Therefore, models that accurately locate the
development and progression of hinging in reinforced concrete bridge systems support
the improvement of calculations for local (damage in hinging regions and beyond) and
global (maximum drift response) performance parameters.
Many researchers have studied the plastic hinge length in column elements
depending on various parameters, such as axial load, yield penetration, shear spreading
of plasticity, shear span-to-depth ratio, and the amount of longitudinal reinforcement.
Various expressions were proposed by several researchers to estimate the plastic hinge
length in these elements. The following are the observations investigated by the previous
researchers:
Axial Load. Extensive research has been conducted to examine the influence
the level of axial load on plastic hinge length (ICC Committee Report (1962),
Park et al. (1982), Priestley and Park (1987), Sakai and Sheikh (1989),
Tanaka and Park (1990), Sheikh (1993, and 1994), Watson and Park (1994),
Kovacic (1995), Bayrak and Sheikh (1997, and 1999), Dodd et al. (2000), and
53
Bae (2005)). The observations, however, are in disagreement. As mentioned

in the studies of Park et al. (1982) and Priestley and Park (1987), the plastic
hinge length was comparatively insensitive to axial load level. Park et al.
(1982) concluded that lp = 0.4h can be used as a simple and safe
approximation for plastic hinge lengths in columns. On the contrary, Tanaka
and Park (1990), Watson and Park (1994), Bayrak and Sheikh (1997, and
1999), Dodd et al. (2000), and Bae (2005) reported that the plastic hinge
length increased with increasing axial load level. Only Watson and Park
(1994) and Bae (2005) proposed expressions to consider the effect of axial
load. Also, Kovacic (1995) justified the ACI (1968) limits for plastic hinge
length for high strength concrete columns with low axial loads. The
expressions are repeated here for clarity:
lp = 0.4*h
(Park et al. 1982)
lc
P
= 1 + 0.4 '
h
f c Ag
P
= 0.3
h P0
lp
As
+ 3
Ag
(Watson and Park 1994)
0.1 L + 0.25 0.25
(Bae 2005)
(1.18a)
(1.20)
Yield Penetration and Shear Spreading. Mander (1983), Priestley and Park
(1987), Paulay and Priestley (1992), and Restrepo (2006) investigated the
effects of yield penetration and shear spreading of plasticity. They proposed
their formulae with respect to the longitudinal reinforcement diameter (db)
and the column height (L or H) based on the experimental results. Paulay and
Priestley (1992), and Restrepo (2006) also considered the effect of yield
stress of the longitudinal reinforcement in their expressions:
54
l p = 0.08 L + 6d b
(Priestley and Park 1987)
(Paulay and Priestley 1992)
(1.16)
(1.17a)
L p = H + f ye d bl 2 f ye d bl
fle
= 1 1 2
=
f
f ye d bl
M ' H2
(
y )
3
My
pH 2
L p = L py +0.06 L
L py = 6.35 d b
l p = 0.05 L + 0.008
f y db
f c'
(Restrepo 2006)
(1.21)
(Mander 1983)
(1.13)
(Berry et al. 2008)
(1.31a)
Shear Span-to-Depth Ratio. Shear span-to-depth ratio was one of the main
variables for Sakai and Sheikh (1989), Tanaka and Park (1990), Bayrak and
Sheikh (1997, and 1999) and Bae (2005). All of the observations given by the
researchers above were consistent that plastic hinge length increased with
increasing shear span-to-depth ratio. Bae (2005) first attempted to study the
relationship between the plastic hinge length and the shear span-to-depth ratio
and proposed a new expression:
P
= 0.3
h P0
lp
As
+ 3
Ag
0.1 L + 0.25 0.25
(Bae 2005)
(1.20)
Member Dimensions. Early expressions for plastic hinge length used the
relationship between the column height, the cross sectional depth, and/or the
diameter of longitudinal reinforcement with the plastic hinge length. The
column height (L) and the depth (d or h) represented the flexural response of
55
the
columns,
whereas
the
diameter
of
longitudinal
reinforcement
corresponded to the slip behavior of the reinforcement. They were not

dimensionless parameters. In later expressions, however, dimensionless
parameters such as axial load ratio, and shear-span-to-depth ratio were
accepted. But, in more recent expressions (Berry et al. (2008)), the member
height (L) and the yield penetration were considered one again to best
represent member behavior.
Dynamic Tests. The studies conducted by Dodd et al. (2000), and Hachem et
al. (2003) utilized the equation proposed by Priestley et al. (1992) to estimate
the plastic hinge length before testing the circular columns under earthquake
loading . This equation was developed using the static test results of 20 bridge
columns having a shear-span-to-depth ratio between 2 and 5.5. According to
dynamic test results, it was concluded that the estimated plastic hinge length
values based on static tests overestimated the measured plastic hinge length
for some columns, but were within a reasonable range of the measured
dynamic test results for other columns. This deserves further study, and will
be investigated in this research.
The expressions that have been developed in the literature are based on
component tests and analyses of their data. The question that remains is how these
estimations of plastic hinge length, damage locations, and estimates of column drift may
change in large scale system tests. This question is the focus of the current study.
56
1.6
OBJECTIVE AND SCOPE
The objective of the current research is to study the plastic hinge length, the
effect of the spread of plasticity on estimates of drift, and the progression of damage in
reinforced concrete bridge systems during response to earthquake motions. The scope of
research to accomplish these objectives includes the followings:
A model is built to estimate the plastic hinge length in reinforced concrete

bridge systems based on experimental results of a large-scale four-span
bridge test. Detailed data along the hinging regions of a reinforced concrete
bridge column are collected during the test to calibrate the model.
Based on an analysis of available test results, a new expression for Lp/d is

proposed.
Parametric analysis is performed to evaluate the effect of axial load, the

amount of the longitudinal reinforcement, shear span-to-depth ratio of the
column, and the slip of reinforcement on the response of the model.
Based on this analysis, a new expression is proposed for plastic hinge length.
Finally, an analytical approach for the spread of plasticity and simple drift
calculations are also developed.
57
PHOTOGRAMMETRIC MEASUREMENTS OF CONCRETE

COLUMN DEFORMATIONS
Equation Chapter 3 Section 1
2.1
INTRODUCTION
The research presented in this document is part of a George E. Brown, Jr.

Network for Earthquake Engineering Simulation (NEES) payload study to investigate
the seismic performance of four-span large scale bridges using conventional and
innovative materials.
The focus of the work is to accurately record and analyze
deformations in plastic hinging locations in concrete bridge structures through an

investigation of actual hinging behavior in large-scale tests of bridge systems subjected
to multiple excitations.
In order to evaluate the plastic hinging regions, a
photogrammetric method was used to remotely track deformations of the concrete

surface in the joint regions. Photogrammetry is a non-invasive technique of remote
visualization of the target components and a computer rendering of the motion (Jauregui
et al. 2006). This computer reconstruction is based on the tracked motion of the target
components. The method of application in this study holds several advantages over
traditional sensors, including experimental components that are simple to construct and
more economical for typical laboratory environments.
Two test specimens were evaluated at the NEES Equipment Site at the University
of Nevada (UNR) in the study. The first was a proof-of-concept test of a bridge column
in which deformations were recorded in one target area.
In the second test, the
photogrammetry method was used to track deformations in the hinging regions of one
bridge column in a large-scale four-span bridge test. Both tests employed strong ground
motion to excite the structure, with the first test in one direction and the second with bidirectional loading.
58
Analysis of deformations in the hinging regions was completed by considering

movement of discrete points on the column surface. This chapter details the specimen
dimensions and material properties as well as the experimental setup to complete the
proof-of-concept test and large-scale bridge test using photogrammetry methods.
Challenges in determining surface deformations are examined, including the resolution
of unclear images and the consistent and predictable definition of points on the surface.
Results for displacement calculations on a gridded surface and cross-section rotation
calculations are discussed.
2.2
PROOF-OF-CONCEPT TEST
On December 14, 2005, a proof-of-concept test for photogrammetry
measurements of concrete elements under earthquake loadings was completed at the

UNR earthquake laboratory. The purpose of this proof-of-concept test was to investigate
the ability of the method to accurately trace the dynamic movement and rotation of the
column as a function of time.
A circular reinforced concrete column (SVTN) was tested under different
earthquake loadings on one of the three biaxial shaking tables in the UNR lab. The
concrete compressive strength was 4.03 ksi (28 MPa) and the yield strength of the steel
in the column was 71 ksi (490 MPa). The diameter of the column was 12 in. (305 mm),
and the total height was 98.5 in (2500 mm). The circular column had 3% longitudinal
reinforcement and 1.82% transverse reinforcement ratios. Figure 2.1 shows the crosssection of the tested column.
59
12 in. [305 mm]

17-#4
Spiral Reinf
0.25 in.
clearcover=0.5 in. [13 mm]
Fig. 2.1 Cross-section of the column
In order to keep track of the displacements on the column surface during the test,
a grid system was applied to the column surface (Fig. 2.2). Figure 2.2 also shows the
traditional linear vertical displacement transducers (LVDT) instrument that was used
during the test. The grid was applied to the surface using a cardboard stencil at the
University of Kansas and black and white paint. First a uniform black background was
applied, and then the stencil was used to apply the white grid marks. The grid consisted
of vertical and horizontal lines spaced irregularly around the face of the circular column.
The irregular spacing of the grid marks allows for a detailed analysis of the column
circular surface. The grid system was applied above the column plastic hinging region to
avoid the congestion of LVDT instrumentation. Displacement and rotation calculations
are reported for the location marked A in Fig. 2.2.
60
Load Direction
Mass
Rig
35.75 in. [908.05 mm]
24 in. [609.6 mm]

4 in. [102 mm]
Fig. 2.2 Grid Setup

The experimental set-up was completed with a camera tower, cameras and
recording system. The specific equipment and materials used in this test are listed
below:
A system of two cameras: model name DXB-9212EF and 0.5-in. (13 mm)
Starlight 600 TVL super high resolution black-white recording properties.
The angles from left and right cameras to the column were 60 (Fig. 2.3).
Video-lens zooms: 60-300mm (2.36 in. 11.81 in.) F4-5.6 capturing

capability.
VCR systems: JVC HR-S2902U brand and high resolution stereo video
cassette recorder.
Monitor: a black and white monitor to show the field of view of the
camera.
An aluminum tower:
an aluminum frame made with 2-in. (51 mm)
thickness extruded aluminum material. The tower has 2 levels to place

61
cameras. Only the first level of the tower used in the proof-of-concept
test. Because the height of the first level was 94.5 in. (2400 mm), metal
and masonry units were placed under the legs of the tower to enable the
cameras to photograph at the level of the grid (Fig. 2.4).
Connections: small aluminum connections which are 30, 45 and 60angle pieces.
86
3.
10
m
]
]3
6
in
.
60
RC
67
.8
6
in
.[
1.
72
m]
50.91 in. [1.29 m]
146.8 in. [3.73 m]
60
SVTN
n.
[2.
24
88
i
60
64
2.
.[
LC
[0
.9
1
m]
in
88
[
in.
4
2.2
Shake Table
LC Location of Left Camera

RC Location of Right Camera
Fig. 2.3 Location of the column and the tower position
62
Tower
Fig. 2.4 The aluminum tower setup

The earthquake motion applied during the tests was recorded at the Rinaldi
receiving station during the 1994 Northridge earthquake (6.7 Richter magnitude). The
maximum acceleration of the record is 0.838g. The Rinaldi record was scaled to obtain
earthquake motions with peak ground accelerations from 0.15g to 1.66g.
acceleration history for the original record is shown in Fig. 2.5.
63
The
Acceleration History-Rinaldi (RRS)

0.8
0.6
Acceleration (g)
0.4
0.2
0
0
10
15
-0.2
-0.4
-0.6
-0.8
Time (sec.)
Fig. 2.5 Acceleration history of the Rinaldi earthquake record
Due to setup limitations in the lab, the distance between the cameras and the
column was 88 in. At this distance, the field of view was too narrow to include two
vertical lines in the photographs at all times. Therefore, the proof-of-concept test focused
on vertical rotation of the column and did not consider the cross-sectional rotation of the
column.
Calculated
displacements
and
rotations
using
the
LVDT
data
and
photogrammetry measurements are compared at Point A on the painted grid (Fig. 2.2).
Figure 2.6 shows the comparison for calculated displacements. Because of a limitation
of the number of frames that may be considered in a single analysis, the photogrammetry
results are shown in two pieces (Run1L and Run2L). The results from two runs are
superimposed and demonstrate good continuity for calculated displacements.
The LVDT displacement values are obtained by first considering the relative
displacement between the displacement transducer recordings at the top of the column
and the absolute displacement of the shaking table.
64
Displacements at Point A are
estimated by linearly interpolating between the relative displacement at the top of the
column and zero displacement at the base to the height of Point A. Displacements values
using the photogrammetry data are determined as the relative motion of Point A. As
shown in Fig. 2.6, the displacement results in terms of periodicity and amplitude of
motion are well-represented by the photogrammetry analysis. The permanent
deformation noted at the end of the test is also captured in the photogrammetry analysis.
Rotations of the horizontal line are compared in Fig. 2.7.
Rotations of the
horizontal line Cross section rotations using the LVDT data are first estimated at each
LVDT location by dividing the relative vertical displacement between two opposing
LVDTs by the column width. As expected, the change in cross section rotation over the
column height of the hinging region decreased with increasing column height.
To
estimate the rotation at Point A, it was assumed that the change in rotation between the
two highest LVDT locations remained constant to Point A.
Rotations of the cross-section based on photogrammetry measurements were
estimated assuming no shear distortions occurred in the grid region.
Using this
assumption, a cross-section rotation may be estimated by considering the vertical

rotation between two points. An average rotation over the height of the column between
the base and Point A was then calculated as the transverse displacement of Point A
divided by the height to that point. In the comparison of vertical rotation calculations
(Fig. 2.7), the periodicity of the rotational response is again well-represented, but
amplitudes are as much as 13% less than the LVDT values. This difference is likely
because of the rough estimation that is provided by extrapolated rotations from LVDT
data and average rotations provided by the photogrammetry measurements.
65
3.0
2.5
2.0
X-Displacement (in.)
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
0
10
11
12
13
14
15
16
17
18
19
20
Time (sec.)
LVDT
Run 1L
Run 2L
Fig. 2.6 Displacement @ Grid Level A vs. Time (Rinaldi 0.95g) (1 in. = 254 mm)
4.0
3.5
3.0
2.5
Rotation (degrees)
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
0
10
11
12
13
14
15
16
Time (sec.)
LVDT
Run 1L
Run 2L
Fig. 2.7 Rotation @ Grid Level A vs. Time (Rinaldi 0.95g)
66
17
18
19
20
2.3
LARGE SCALE FOUR-SPAN RC BRIDGE TEST

A four-span large scale reinforced concrete bridge was tested on February 12-15,
2007 at the UNR laboratory. Representatives from NEES, CALTRANS, UNR, KU,
other participating universities, and local media congregated in the test area.
Photogrammetry measurements were taken in two hinging regions, which required the
use of four cameras and two grid configurations. The significance of this test was that it
was a large-scale bridge with columns having double curvature response and resisting
massive earthquake forces. Also, biaxial earthquake loading was employed. Challenges
that are addressed in the photogrammetry analysis included irregular image quality and
inconsistent definition of points on the concrete surface.
2.3.1
Description of Specimen
The test specimen consisted of a four-span reinforced concrete bridge with end
abutments as shown in (Fig. 2.8). The bridge was quarter-scale with two interior spans
that were 29 ft. (8.84 m) in length, and two exterior spans of 24.5 ft. (7.47 m) for a total
length of approximately 110 ft. (33.5 m). The clear heights of the bents were 5, 6 and 7
ft. (1.52, 1.83, and 2.13 m), with the tallest bent in the middle (Bent 2). The
superstructure consisted of a solid slab that was post-tensioned in both the longitudinal
and transverse directions. The bridge also included abutment seats at both ends that were
driven in the longitudinal direction by dynamic hydraulic actuators. The depth of the cap
beams was 15 in. (0.38 m). The total length of the cap beams (perpendicular to the
longitudinal axis of the bridge) was 98 in. (2.49 m). The concrete compressive strength
used in the bridge (excluding the slab) was 6.7 ksi (46 MPa) and reinforcement was
ASTM A706 Grade 60. The bents were numbered Bent 1, Bent 2, and Bent 3 starting
from left (South) to right (North). The east column of Bent 3 was selected to be the target
67
structure for applying the painted grids. Figures 2.9, 2.10 and 2.11 show the elevation
and side views of Bents 1, 2 and 3, respectively.
N
282 in. [7.16 m]
348 in. [8.84 m]
348 in. [8.84 m]
282 in. [7.16 m]

14 in. [0.36 m]
SHAKE TABLE
Abutment 1
SHAKE TABLE
Bent 1
SHAKE TABLE
Bent 2
Bent 3
Abutment 2
Fig. 2.8 Elevation view of the four-span bridge
15in [0.38m]
98in [2.49m]
60in [1.52m]
1.5in [0.04m]
5.5in [140mm]
63in [1.6m]
12in [305mm]
12in [305mm]
5.5in [140mm]
15in [381mm]
12in [305mm]
18in [457mm]
36in [0.91m]
108in [2.74m]
Fig. 2.9 Elevation and side view of the Bent1
68
1 5 in [ 3 8 1 m m ]
9 8 in [2 .4 9 m ]
6 3 in [1 .6 m ]
1.5in [38mm]
5.5in [140mm]
12in [305mm]
12in [305mm]
5.5in [140mm]
1 5 in [3 8 1 m m ]
1 2 in [3 0 5 m m ]
8 4 in [ 2 . 1 3 m ]
1 8 in [4 5 7 m m ]
1 0 8 in [2 .7 4 m ]
3 6 in [ 0 . 9 1 m ]
Fig. 2.10 Elevation and side view of the Bent2
98in [2.49m]
1 5in [0 .3 8m ]
1.5in [0.04m]
5.5in [140mm]
63in [1.6m]
12in [305mm]
12in [305mm]
5.5in [140mm]
15in [381mm]
7 2in [1.83 m ]
1 2in [30 5m m ]
3 0in [0 .7 6m ]
36 in [0.91 m ]
108in [2.74m]
Fig. 2.11 Elevation and side view of the Bent 3
2.3.2
Experimental Setup
The top and bottom hinging regions of the east column of Bent 3 were the focus
for collecting photogrammetry data in the large-scale four-span reinforced concrete

bridge test. Two grid systems were applied (Fig. 2.12). The bottom grid consisted of
vertical and horizontal lines spaced irregularly around the face of the circular column
(Fig. 2.13). As shown in Fig. 2.4, the same grid system was applied in the proof-of69
concept test. The top grid used a different configuration consisting of lines and squares
(Fig. 2.3b) that resembled the spacing of the lines in the bottom grid. The top grid was
applied to track the displacements of the isolated points (squares), similar to applications
in earlier photogrammetry methods that were completed at the University of Kansas
laboratory (Reetz, 2005).
To apply the grid systems, the column was first spray-painted black for a solid
dark background. The lines and squares were lightly traced with a pencil, and then a
single stencil was used to fill in the lines and squares using a white paint.
The same aluminum tower as described in the proof-of-concept test Section (2.2)
was used to support the cameras, including the second level of the tower for the tests.
Two cameras were mounted on each level (at 94.5 in. (2400 mm) and 159 in. (4039 mm)
height) (Fig. 2.14). In the first level of the tower, due to changes in the height level of
the hinge region, the height of the cameras was increased to directly view the bottom
grid region. This was accomplished by placing small aluminum columns to mount the
cameras on the tower (Fig. 2.14). Metal weight pieces were placed at each side of the
base of the tower to provide balance on the strong floor (Fig. 2.15). There were four
VCR systems and two monitors (two channels for each monitor) used in the test.
70
Fig. 2.12 Grid systems on the Bent3 east column surface
(a)
(b)
Fig. 2.13 Close view of (a) bottom and (b) top grid systems
71
Fig. 2.14 The aluminum tower and four cameras
Fig. 2.15 Metal weight pieces placed on each side of the tower
72
The distance between the east column of Bent 3 and the tower was 288 in. (7315
mm). The angle from the left camera to the column was 73.58, and from the right
camera to the column was 44.97 on the horizontal plane (Fig. 2.16). The tower was
located on the north-east side of the UNR laboratory.
Figure 2.17 shows the numbers of all points defined by the grid systems. The
points on the bottom and top grid systems were defined at the intersections of the thick
vertical and horizontal lines. The locations of a series of linear vertical displacement
transducers (LVDTs) and one displacement transducer (DT7) monitored by the research
team at UNR are also shown in Fig.2.17.
m]
6.
75.24
14
n.
8i
.9 1
50 ]
3m
3 .7
in .
9
1 .2
m]
6 7 .8
RC
75.
6 in .
[ 1 .7 2
m]
24
Tower
288
LC
1 0 3 .8
6 in.
[2.64
m]
3 6 in
. [0 .9
1
.32m
in. [7
28
8
.[
7.
32
.5 3
in
29
m
]
Bent 3
Shake Table
Fig. 2.16 Location of the Bent 3 east column and the aluminum tower
73
DT7
3ETR3
57
56 58
50
49 51
59
52
62
6061
55
5354
43
42 44
45
48
4647
3ETR5
3ETR6
36
35 37
29
28 30
212223
41
3940
34
31 3233
24 252627
38
1617 18
19 20
1112
13
1415
6 7
9 10
1 2
4 5
3EBR5
3EBR7
3ETR4
3EBR6
3EBR8
Fig. 2.17 Grid system and LVDT locations on column in the Bent 3 east column
2.3.3
Earthquake Loading
The earthquake motion used in the tests consisted of biaxial and uniaxial
applications of scaled motions measured at the Century City Country Club during the
1994 Northridge, California earthquake. This motion was applied 13 times with six
different scaled intensities having increased amplitude so that the progression of damage
could be tracked from pre-yield to failure as shown in Table 2.1.
A white noise
excitation was applied before each change in ground motion intensity to test for system
identification.
The first damage to the bridge system occurred in the columns of Bent 3 (the bent
with medium height) and then the columns of Bent 1 (the bent with the shortest height).
74
The earthquake intensity was not increased after the 1.20g motion (Test 6) because the
reinforcement in the left column of Bent 1 started to buckle.
Almost all of the
longitudinal bars in the columns of Bent 1 had buckled by Test 6. The same motion was
repeated one more time before the test was completed.
Other damage that was noted during the test included small amounts of cracking
in the bent caps, and some spalling of the cover concrete of the abutment seats and the
superstructure. The cracks at the bent caps began after the first test (Test 1A) and
progressed until the tests were done. The spalling of the concrete cover of the abutment
seats occurred after Test 6.
75
Table 2.1 Earthquake motion levels applied during the test

Test
No.
WN01
WN02
1A
1B
1C
1D
WN11
WN12
2
WN21
WN22
3
WN31
WN32
4A
4B
4C
4D
WN41
WN42
5
WN51
WN52
6
WN61
WN62
6
2.3.4
Test Date
2/12/2007
Motion
Level
1
1
1
1
2/13/2007
4
4
4
4
2/14/2007
5
2/15/2007
Motion PGA (g)

Trans.
Long.
White Noise (Trans.)
White Noise (Long.)
W/Restrainer1
0.09
W/Restrainer2
0.09
Longitudinal
0.09
Biaxial
0.075
0.09
White Noise (Long.)
Biaxial
0.15
0.18
White Noise (Long.)
Biaxial
White Noise (Long.)
W/Restrainer1
0.6
W/Restrainer2
0.6
Longitudinal
0.6
Biaxial
0.5
0.6
White Noise (Long.)
Biaxial
0.75
0.9
White Noise (Long.)
Biaxial
1
1.2
White Noise (Long.)
Biaxial
1
1.2
Test Type
Results
Due to shaking effects on the strong floor during the test, there is a chance for
movement of the base of the aluminum tower. In order to eliminate the extra movement
76
from the tower shakings, a fixed point on the wall was identified and tracked during the
strong excitations (Fig. 2.18).
Fig. 2.18 Fix point on the wall at the back of the column
The lateral displacement of the fixed point on the wall (Fig. 2.19) was determined
and subtracted from the lateral displacements of the grid points to obtain the
displacements of the column caused by the shake table motions. The calculated tower
displacements as indicated by the movement of the fix point are very close to zero in
value (within one or two pixels of value) and did not significantly affect the results. The
vertical displacement of the fixed point was also calculated (Fig. 2.20), but it does not
need to be subtracted from the vertical displacement of the grid point because crosssectional rotation calculations are based on relative vertical movement between two
points on the column.
77
Lateral Displacement of Fix Point

X Displacements (in.)
0.1
0.05
0
0
10
15
20
25
30
35
-0.05
-0.1
Time (sec.)
Fig. 2.19 Lateral displacement of fixed point on the wall at Test 4D (1 in. = 254 mm)
Vertical Displacement of Fix Point

Y Displacements (in.)
0.015
0.01
0.005
0
-0.005 0
10
15
20
25
30
35
-0.01
-0.015
Time (sec.)
Fig. 2.20 Vertical displacement of fixed point on the wall at Test 4D (1 in. = 254 mm)
2.3.4.1 Definition of Points on Surface

Deformations were evaluated considering displacements and cross-sectional
rotation of various locations on the column. For each deformation that was calculated, it
was necessary to have a consistent and predictable definition of points on the column
surface. Challenges in the photogrammetry analysis included some instances of poor
images and the reliability of the selected method to define the points, as described below.
78
There are some differences in the data provided by LVDT instrumentation and
that recorded for photogrammetry application. The rotation calculations obtained using
photogrammetry analysis represents the cross-sectional rotations based on the movement
of the column, which includes not only flexural but also rigid body rotation. Rotations
calculated using the LVDT data from the test includes only flexural contributions. The
LVDT instrumentation was set to record movement in the transverse and longitudinal
planes, and photogrammetry measuruments were recorded at an angle between the
LVDT instrumentation. This difference in orientation is accounted for in the analysis
and results presented in this section.
Unclear Images
Interactive Data Language (IDL) was used to analyze the video images that were
recorded during the test. Each video was first divided into over one thousand images for
individual analysis. The displacement of a unique point in an image was defined as the
relative horizontal movement of that point between successive images. However, at
times it was difficult to keep track of a unique point because of the quality of the images.
The image (640 x 480) was written as 480 lines that are 640 pixels wide. The
odd-numbered lines are written first, followed by the even-numbered lines so that the
image is really two images offset by 1/60th of a second. The first partial image (half of
the total image) is comprised of the odd lines, and the second partial image is
comprised of the even lines. Due to the shaking effect from the strong floor during the
tests and the incompatibility in recording frequency between the cameras and the
earthquake motions, many images were compromised. Some of the images appeared as
double lines for each horizontal and vertical line, and some were simply blurry images.
Table 2.2 shows the number of corrupted images that were recorded during each test
79
motion and the total percentage of corrupted images recorded by each camera. The
Image Group number in the table signifies a grouping of images that could be analyzed
at one time. Therefore, Image Group 1 contains images from the earliest time in a test.
Each image represents 1/30th of a second during the test. The total number of images
reviewed in a single analysis ranged from 339 to 998. The percentage of corrupted
images ranges from 0-62%. For recordings at the top of the column, the percentage of
corrupted images exceeded 4% when at least one component of shaking exceeded 0.5g
with a maximum of 30% corrupted images (Test 4D Top Right Camera). Recording
quality was much more sensitive at the base of the column, with as much as 24%
corrupted images during Test 2 (Bottom Left Camera), which had a peak ground
acceleration of only 0.15g. The images recorded early (Image Group 1) in each shake
table test ranged from 8-62% corrupted, the recordings from Test 5 Bottom Left Camera
having 62% corrupted images.
80
Table 2.2 Corrupted images during the tests

Camera
(Fig. 2.16)
Test
Top Left
2
Top Right
2
Bottom Left
2
Bottom Right
2
Image
Groups
No. of
corrupted
images
27
28
80
72
No. of total
images
899
299
899
349
339
339
339
339
339
339
Corruption
Percentage
24
21
Test
4D
4D
4D
4D
Image
Groups
No. of
corrupted
images
100
102
10
171
40
193
10
No. of total
images
339
339
339
339
339
339
339
339
339
339
339
339
Corruption
Percentage
29
30
50
12
57
Test
Image
Groups
No. of
corrupted
images
220
136
210
55
180
15
No. of total
images
899
899
339
339
339
339
339 339
Corruption
Percentage
24
15
62
16
53
Test
Image
Groups
No. of
corrupted
images
210
200
221
264
No. of total
images
999
339
999
339
999
339
999
339
Corruption
Percentage
21
20
22
26
81
One option to improve calculations of displacements and rotations using the

corrupted images is to consider only the even-numbered or odd-numbered lines,
separately. If the even and odd lines are separated, the double image effect that is
produced by a shift between the two halves of the image being written can be avoided by
considering only one half image at a time. Figures 2.21 and 2.22 show the results of
the analysis of even lines, odd lines and the combination of the two for the entire test 4D
and a 5 second interval for each analysis, respectively. It is difficult to distinguish the
results in Fig. 2.21. As shown in Fig. 2.22, the even- and odd-line analyses are very
similar in smoothness, but the combined image shows irregularities in the movement of
the point.
Point 7 Vertical Displacement (Test4D Bottom)
0.25
Even
Odd
Vertical Displacement (in.)
0.20
Comb.
0.15
0.10
0.05
0.00
-0.05
10
15
20
25
30
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.21 Point 7 vertical displacement at the bottom grid system (1 in. = 254 mm)
82
35

0.25
Even
0.20
0.15
0.10
0.05
0.00
5
10
-0.05
-0.10
-0.15
-0.20
Time (sec.)
a) Even Line
0.25
Odd
0.20
0.15
0.10
0.05
0.00
5
10
-0.05
-0.10
-0.15
-0.20
Time (sec.)
b) Odd Line
0.25
Comb.
0.20
0.15
0.10
0.05
0.00
5
10
-0.05
-0.10
-0.15
-0.20
Time (sec.)
c) Combination
Fig. 2.22 Five second interval to compare even- odd- and combination-line analyses (1
in. = 254 mm)
83
The vertical displacement results of Point 7 in the even and odd line analyses are
compared with the LVDT 3EBR7 reading (the closest UNR measurement to the point on
the bottom grid system as shown in Fig. 2.17) in Figs. 2.23-2.24. Figure 2.25 shows only
the LVDT 3EBR7 recorded displacements for clarity. Because of the larger disparity
noted in the figures between peak displacement values recorded in the odd-line analysis
(Fig. 2.24) and LVDT recordings, the even line approach (Fig. 2.23) is selected to
provide better photogrammetry results.
Rotation of horizontal line is calculated as the relative vertical displacements of
the horizontal line that Points 7 and 8 align and divided by the horizontal distance
between the points. For the photogrammetry analysis, the horizontal distance between
the points changes within each image. Figure 2.26 shows the rotation of horizontal line
calculated using Points 7 and 8 at 12 inches (305 mm) from the bottom fixity in the
transverse direction. The appearance of noise with nearly constant magnitude is a
result of single pixel differences in calculated displacements (pixelization).
0.25
Even
LVDT-3EBR7
0.20
0.15
0.10
0.05
0.00
-0.05
10
15
20
25
30
35
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.23 Comparison of Point 7 vertical displacement for even lines with LVDT 3EBR7
data (1 in. = 254 mm)
84

0.25
Odd
0.20
LVDT-3EBR7
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.24 Comparison of Point 7 vertical displacement for odd lines with LVDT 3EBR7
data (1 in. = 254 mm)
LVDT 3EBR7 (Test 4D)
0.25
0.20
0.15
0.10
0.05
0.00
-0.05 0
-0.10
10
15
20
25
30
-0.15
-0.20
Time (sec.)
Fig. 2.25 LVDT 3EBR7 vertical displacement history (1 in. = 254 mm)
85
35
0.030
0.025
0.020
0.015
Rotation (rad)
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
-0.015
-0.020
-0.025
Time (sec.)
Fig. 2.26 Rotation of horizontal line calculated from Points 7 and 8 at the bottom grid in
the transverse direction
Intersecting Lines to Define Points
There are several options for defining each unique point on the grid, all of which
were considered and the best overall representation selected as described in this section.
Points on the grid are defined by the intersection of the thick vertical and horizontal
lines. The vertical and horizontal lines are 0.25 in. (6 mm) thick; therefore, there are
several ways to define a line within that thickness. One option is to seek a definition of
the mid-point of each 0.25-in. (6 mm) thickness and define a middle line as the
connection of those points (Fig. 2.27). Another option is to define the edges of the lines.
For this option, the edge lines may be defined by points of a particular light intensity
value, or a change in relative intensity between the black to white paint transition.
86
Finally, the length of the line that is used to define each Point on the grid also will
influence results. Each of these options is discussed below.
Middle Line
Edge Lines
Edge Line Point 3
Middle Line Point 3
Fig. 2.27 Close up of lines used to define Point 3 (Fig. 2.17)
1. Middle Lines: In IDL program, the intensity of each pixel was determined
for each image in the analysis. To define the middle lines, a threshold value
of light intensity (Robert threshold) was used to define the left and right
boundaries for vertical lines and top and bottom boundaries for horizontal
lines in the images. The definitions of the boundaries of the lines were
completed for a cropped image area. The intensity of each pixel of a given
line was then determined between the defined boundaries for the vertical and
horizontal lines. In theory, the maximum pixel intensity would define a
unique line through the middle portion of each 0.25-in. (6 mm) line. The
locations of the maximum pixel value along the thickness of the lines near
Point 8 are shown in Fig. 2.28. To get refined results for the maximum
intensity pixel locations, the area around the maximum intensity pixel
locations were searched in detail. Because the location of the maximum
pixel intensity varied over the length of each line, the middle points did not
87
align to form a smooth middle line. Therefore, the displacements and the
cross-sectional rotation results obtained using this method were not very
accurate.
The vertical displacements of Point 7 and 9 are compared with LVDT
3EBR7 and 3EBR8 readings in Figs. 2.29 and 2.30.
As shown in the
figures, the peak values represent LVDT measurement fairly well, but there
is a lot of high frequency noise in the plots. The resulting calculated rotation
at about 12 in. (305 mm) from the footing of Bent3 is shown in Fig. 2.31.
As expected, the detailed calculation of the rotation varies widely.
225
133 180 214
211 182 142 108 90
219
102 149 189 214
203 170 127 77
0.25 in. [0.01 m]

Fig. 2.28 Pixel intensities of two different levels in an image
88
Point 7 Vertical Displacement (Test 4D Bottom)

0.30
Middle Line
LVDT-3EBR7
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
Time (sec.)
254 mm)
0.30
Middle Line
LVDT-3EBR8
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
Time (sec.)
254 mm)
89
0.030
0.025
0.020
Rotation (rad)
0.015
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
Time (sec.)
Fig. 2.31 Rotation of horizontal line using Points 7 and 9
To enhance the middle-line calculations, an average cross-sectional rotation over

the height defined between two vertically connected points was calculated. For example,
the vertical and horizontal displacements of Points 7, 8, 12 and 13 were first determined
using the average vertical displacements of Points 7 and 12 and Points 8 and 13 were
then calculated. An average rotation between these points was determined and is shown
in Fig. 2.32. The average rotation obtained from IDL results at about 10 in. (254 mm)
from the bottom hinging level was not improved using this method, with the magnitude
of maximum rotation greatly magnified using the average values.
90
0.15
IDL_h=7.6 - 13.2 in.
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
30
35
40
-0.05
Time (sec.)
Fig. 2.32 Average rotation of four Points on the column surface
2. Edge Lines: Because of the thickness of the horizontal and vertical lines that
define the grid system, the exact unique location of each Point (generally
located at the intersection of the thick horizontal and vertical grid lines) may
be defined in several ways.
The middle-line method, as discussed
previously, did not yield reliable results. Another method is to consider the
edges of the thick grid lines to define two new lines (edge-lines) for each
grid line. Therefore, at the intersection of the thick grid lines that generally
defines a Point, there are actually four intersections of the edge lines to
consider as possible unique locations of that Point (see Fig. 2.33, for
example, Point 12 and Point 13 a, b, c, and d).
In IDL programming
language, the equations of the edge lines were obtained by defining the
91
boundaries of the thick grid lines using the constant Robert threshold value,
as described in the previous section.
Possible corner locations of the
selected points (locations a, b, c, or d in Fig. 2.33) were calculated from the

intersections of the four edge lines. Cross-sectional rotations were then
computed using the same corners of paired Points (for example, a, b, c and d
matching corners of Points 12 and 13).
Figures 2.34 through 2.37 show the rotation of horizontal line results
calculated between the corners of the Points.
Also, these four corner
rotations were averaged to obtain the middle point (Point 12 and 13)
rotation response. Figure 2.38 shows the middle point rotation result for
Test 4D. The middle Point was selected to calculate the rotations for
photogrammetry results because the corner rotation results give high values
and more high frequency noise than the middle point rotation result.
0.25 in. [0.01 m]

a
0.25 in. [0.01 m]
0.25 in. [0.01 m]

a
0.25 in. [0.01 m]
13
12
c
~2.5 in. [0.06 m]
Fig. 2.33 Four corners surrounding general location of Point 12 and 13
92
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
30
35
30
35
-0.05
-0.10
Time (sec.)
Fig. 2.34 Corner (a) rotation of Point 12 and 13
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
-0.05
-0.10
Time (sec.)
Fig. 2.35 Corner (b) rotation of Point 12 and 13
93
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
30
35
30
35
-0.05
-0.10
Time (sec.)
Fig. 2.36 Corner (c) rotation of Point 12 and 13
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
-0.05
-0.10
Time (sec.)
Fig. 2.37 Corner (d) rotation of Point 12 and 13
94
0.08
Rotation (rad)
0.06
0.04
0.02
0.00
0
10
15
20
25
30
35
-0.02
-0.04
Time (sec.)
Fig. 2.38 Average rotation of Point 12 and 13
3. Selecting Points to Define Lines- Constant Robert Threshold vs.

Intensity Difference: Two different approaches were used to define the
boundaries of the edge lines.
In the first approach, a constant Robert
threshold value that adequately defined the transition from black to white
paint was determined from the first image in each image group. A line
length was assigned and the equations of the four edge lines were calculated.
Various degrees of polynomials were considered to define the edge lines, but
simple first degree (linear) representation proved to be the best fit. The edge
line lengths were set between values of 0.5 in. (13 mm) and 1.0 in. (25 mm)
depending on the relative distance between the grid lines. Even-numbered
lines were used in the analysis, as described previously (Section 2.3.4.1
Unclear Images).
95
The vertical displacements of Point 7 and 9 were tracked during Test 4D

and compared with LVDT 3EBR7 and 3EBR8 readings (Figs. 2.39-40). The
vertical displacement results for Point 7 and 9 follow the general periodicity
and have similar peak values as the LVDT measurements. The rotation of
horizontal line obtained using the constant Robert threshold approach is
shown in Fig. 2.41. As shown in Fig. 2.41, due to pixelization there is some
constant noise at the end of the test.
In the second approach, the intensity differences in consecutive pixels
were obtained using the IDL program, and the maximum difference was
used to define the boundaries for the four edge lines. All other parameters
remained constant.
Figures 2.42-2.43 show the vertical displacements of Point 7 and 9 as
compared with LVDT measurements.
Rotation of horizontal line was
computed using the relative vertical displacements of the two points

normalized to the horizontal distance between the points in Fig. 2.44. As
shown in Figs. 2.39 through 2.44, using the constant Robert threshold
method produces more reliable results. The intensity-difference approach
results in approximately twice the magnitude of the rotation than the
constant Robert threshold method, and much more high frequency noise in
displacements and the rotations.
96

0.25
Robert Threshold
LVDT_3EBR7
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
Fig. 2.39 Point 7 vertical displacement vs. LVDT 3EBR7 data (1 in. = 254 mm)
V e rtic a l D is p la c e m e n t (in .)
0.25
Robert Threshold
LVDT-3EBR8
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
Fig. 2.40 Point 9 vertical displacement vs. LVDT 3EBR8 data (1 in. = 254 mm)
97
0.030
0.025
0.020
Rotation (rad)
0.015
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
Time (sec.)
Fig. 2.41 Rotation of horizontal line obtained using the constant Robert threshold
0.30
Intensity Difference
Vertical D isplacement (in.)
0.25
LVDT-3EBR7
0.20
0.15
0.10
0.05
0.00
-0.05
10
15
20
25
30
35
-0.10
-0.15
-0.20
Time (sec.)
mm)
98

0.30
Intensity Difference
LVDT-3EBR8
0.20
0.10
0.00
0
10
15
20
25
30
35
-0.10
-0.20
Time (sec.)
mm)
0.06
Rotation (rad)
0.04
0.02
0.00
0
10
15
20
25
30
-0.02
-0.04
-0.06
Time (sec.)
Fig. 2.44 Rotation of horizontal line at h=7.7 in. (196 mm) from the bottom fixity
99
35
4. Long Lines vs. Short Lines: When using the constant Robert
threshold method, the length of the edge line controls the number of points
used to define that line, and therefore, the equation of the line. Longer (1.5
in. (38 mm) to 3.0 in. (76 mm)) edge-line lengths were investigated to
determine the influence on the definition of a unique Point. Even-numbered
lines were used in the analysis.
The vertical displacements of Point 7 and 9 were determined and
compared with LVDT 3EBR7 and 3EBR8 readings (Figs. 2.45-2.46). The
resulting cross-sectional rotation is shown in Fig. 2.47 and compared with
the calculated rotation from LVDT data at approximately 12 in. from the
footing of Bent3. The short line and long line analyses can be compared by
examining cross-sectional rotation calculated from each analysis (Figs. 2.41
and 2.47, respectively). There is approximately 0.011 radians of constant
noise shown in the long-line results (Fig. 2.47), whereas the short-line
results (Fig. 2.41) show approximately half the magnitude and less
frequency of noise. The short-line definition was selected for all ensuing
calculations.
100

0.25
Long-Line
0.20
LVDT-3EBR7
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
254 mm)
0.25
Long-Line
V e rtic a l D is p la c e m e n t (in .)
0.20
LVDT-3EBR8
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
254 mm)
101
0.030
0.025
0.020
Rotation (rad)
0.015
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
Time (sec.)
Fig. 2.47 Rotation of horizontal line calculated using long-line
Evaluation of Grid System

As described in Section 2.3.2, two separate grid systems were used to
evaluate the displacements for photogrammetry analysis. In the displacement

calculations of the top grid system, Unique points in the targets (squares) were
defined by seeking the maximum pixel intensity over the thickness of the square,
similar to Middle Lines analysis described in Section 2.3.4.1 Intersecting Lines.
However, the value of pixel intensity did not change significantly over the
thickness of the square, and the definition of the Point location was found to vary
from frame to frame. An example of the calculated displacement of Point 46
defined using the maximum intensity in the square is shown in Fig. 2.48 and
compared with LVDT data. By contrast, results obtained using the Edge Lines to
102
calculate corner points and the Average Point location is shown in Fig. 2.49 and
compared with LVDT data. As shown in both figures, the vertical displacement
of the middle point follows the LVDT 3ETR4 data when the Edge Lines method
is used and gives much closer results than the vertical displacement using the
maximum intensity approach.
The Vertical Displacement (Point 46) -(Test 6 Top)

0.35
Max Intensity
0.30
LVDT-3ETR4
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
-0.25
-0.30
Time (sec.)
Fig. 2.48 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data
using the maximum intensity approach (1 in. = 254 mm)
103
The Vertical Displacement (Point 46) - (Test 6 Top)

0.35
Even-Line
0.30
LVDT-3ETR4
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
-0.25
-0.30
Time (sec.)
Fig. 2.49 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data using
the Edge Lines method (1 in. = 254 mm)
Rotation of Vertical Lines vs. Rotation of Horizontal Lines on Column

Surface
As described in Section 2.2 Proof-of-Concept Test, a rotation of vertical line for

the column can be calculated using transverse movements of points that are aligned
vertically, and this rotation represents the average cross-sectional rotation between those
two points if shear deformations can be neglected.
To see the difference between
calculated rotations of vertical lines and rotations of horizontal lines, movements of

points on the bottom and top grid systems were determined (Fig. 2.17). The rotation of
vertical line was calculated using the relative lateral displacements of two verticallyaligned points on the grid surface divided by the vertical distance between the points.
Rotations of horizontal lines are calculated as described previously.
104
The lateral displacements of Point 3 and 8 on the bottom grid were tracked during
Test 4D motion to calculate rotations of vertical line (Fig. 2.50). Points 7 and 9 shown in
Fig. 2.17 were selected and the vertical movements of the Points were determined to
calculate the rotation of horizontal line during Test 4D. The rotation of horizontal line
results is shown in Fig. 2.51. Also, the rotation of vertical line and the rotation of
horizontal line are compared in Fig. 2.52.
In order to compare the rotation of vertical line and rotation of horizontal line at
the top grid, Point 38 and 52 were selected and tracked during Test 4D. Figure 2.53
shows the rotation of vertical line results calculated between Point 38 and 52. Point 44
and 46 were selected to determine the rotation of horizontal line shown in Fig. 2.54. The
comparison between the rotation of vertical line and rotation of horizontal line results are
shown in Fig. 2.55. As an average measurement, the rotation of vertical line provides a
smoother trace than the rotation of horizontal line calculations from the photogrammetry
data. Comparisons for the three different tests at each cross-sectional rotation location
are found in Appendix A.
105
0.02
0.015
Rotation (rad)
0.01
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
Time (sec.)
Fig. 2.50 Rotation of vertical line calculated using Point 3 and 8
0.02
0.015
0.01
Rotation (rad)
0.005
0
-0.005
10
15
20
25
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
Fig. 2.51 Rotation of horizontal line using Point 7 and 9
106
30
35
0.02
Vertical_Rotation
0.015
Cross-sec. Rot.
0.01
Rotation (rad)
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
Fig. 2.52 Comparison of rotations (local vs. average) in Test 4D
0.04
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec.)
Fig. 2.53 Rotation of vertical line calculated using Point 38 and 52

107
35
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec.)
Fig. 2.54 Rotation of horizontal line using Point 44 and 46
0.05
Vertical_Rotation
0.04
Cross-sec. Rot.
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec.)
Fig. 2.55 Comparison of rotations (local vs. average) in Test 4D
108
35
2.3.4.2 Displacements
The grid that is used to provide photogrammetry data is advantageous when
considering the deformation to the column surface over a continuum, which may be
correlated with detailed Finite Element results. By contrast, LVDT instrumentation will
provide only discrete displacement data in a single direction. In addition, if significant
damage occurs to the concrete surface, LVDTs measurements may become
dysfunctional, but the painted grid surface may be reconstructed in an image and
photogrammetry data preserved. This section describes the displacement calculations in
the gridded surfaces for multiple earthquake excitations. Displacement histories for all
gridded Points and earthquake motions are shown in Appendix B.
Top of the Column
Point 59 in the top grid system (Fig. 2.17), which was the closest point to the top
of the column, was selected to track the peak displacement along the direction that the
left cameras have 15.25 counter clockwise from the longitudinal direction of the bridge.
The displacements at grid Point 59 captured during Test 2, Test 4D, and Test 6 are
compared with the combined components of DT7, which recorded the transverse
displacement of the Bent 3 at the middle of the bridge deck, and DS5 and DS1
measurements, which measured the gap size history and the displacement of the north
abutment in the longitudinal direction of the bridge, during the same motions in Figs.
2.56, 2.57, and 2.58, respectively. There is approximately 23.3 in. (592 mm) between
the transverse beam and Point 59. Because the column is in double curvature during the
test, the lateral displacements at these two heights are assumed to be similar.
109
Considering Figs. 2.56-2.58, the overall periodicity of the two measurements

correlate very well, with differences in magnitude primarily attributed to the height
flexibilities differentials of the transverse beam and column hinging region. In addition,
the photogrammetry measurements capture movement in the column, which may differ
at discrete time from movement of the cap beam that is captured by LVDT
instrumentations.
Some variations in periodicity between the photogrammetry
displacement results and the readings of the displacement transducers are noted (for
example, between the time periods of 12 and 13.5 seconds in Fig. 2.56), and in these
cases the photogrammetry results have been visually confirmed with the associated
images. As shown in all figures, the lateral displacements are reliable and correlate with
the LVDT instrumentation data.
0.8
Photogrammetry
0.6
DT7 & (DS5+DS1) combined

components
0.4
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
-1
Time (sec.)
Fig. 2.56 Comparison of Drift at Point 59 for Test 2 (1 in. = 254 mm)
110
40
2.5
Photogrammetry
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
Time (sec.)
Fig. 2.57 Comparison of Drift at Point 59 for Test 4D (1 in. = 254 mm)
5.0
Photogrammetry
4.0
DT7 & (DS5+DS1)

combined components
Displacement (in.)
3.0
2.0
1.0
0.0
0
10
15
20
25
30
35
-1.0
-2.0
-3.0
Time (sec.)
Fig. 2.58 Comparison of Drift at Point 59 for Test 6 (1 in. = 254 mm)
111
40
The maximum calculated displacements using photogrammetric measurements

for the three different earthquake motions are compared with the UNR maximum
movement recorded by the combined components of DT7, DS5 and DS1. The maximum
displacements calculated in the photogrammetry analyses were normalized by the
maximum combined displacement of DT7, DS5 and DS1 transducers.
Figure 2.59
shows the ratios for the three different motions plotted in terms of peak ground
acceleration (PGA) for Test2, Test 4D, and Test 6. As shown in Fig. 2.59, the ratio
decreases with increasing motion intensity. The ratio ranges from 1.04 in Test 2 to 0.74
in Test 6. In the higher intensity test, much damage was noticed at the cap beam and
more relative movement between the top of the column and the DT7 location is probable.
In addition, it is important to note that the quality of images decreases with increasing
earthquake intensity and incompatibilities in recording frequency between the cameras
and the earthquake motions occurred more frequently.
1.2
1
Ratio
0.8
0.6
0.4
Ratio
0.2
0
0
0.2
0.4
0.6
0.8
1.2
1.4
PGA (g)
Fig. 2.59 Maximum lateral movement ratios between photogrammetry and LVDT results
.
112
Hinging Regions
The displacement calculations are repeated at each of the grid points along the
hinging surface. Three vertical lines, as defined by the grids shown in Fig. 2.17, are
monitored through each test. On the middle line, the selected Points are 3, 8, 13, and 18
at the bottom grid and 38, 45, 52, and 59 at the top grid system. For the left line, Points
2, 7, 12, and 17 are chosen at the bottom and Points 37, 44, 51 and 58 are selected at the
top. For the right line, Points 4, 9, 14, and 19 and 39, 46, 53, and 60 are calculated at the
bottom and top, respectively.
Plots illustrating the calculated deformed shapes in the hinging region for Test 2
are shown in Figs. 2.60, 2.61, and 2.62. The overall deformed shape of the column for
Test 2 at the maximum column drift is shown in Fig. 2.60. The deformations in the
hinging regions are compared in Figs. 2.61 and 2.62 with the associated column images
at the time of maximum column drift. The deformations (Figs. 2.61-2.62) correlate well
with the associated images. For Test 2, little nonlinear deformation is noted.
113
Bent3 East Column

Maximum Deformed
Shape (Test2)
Middle Line
Left Line
80
Right Line
70
Column Height (in.)
60
50
40
30
20
10
0
-6 -4 -2 0 2 4 6
Drift (in.)
Fig. 2.60 Overall deformed shape along the column height at maximum column drift
.
114
Middle Line
Top Grid Deformed Shape (Test2)
Left Line
Right Line
71
69
Column Height (in.)
67
65
63
61
59
-6
-4
-2
Drift (in.)
Bottom Grid Deformed Shape (Test2)
16
14
Column Height (in.)
12
10
4
Middle Line
Left Line
Right Line
0
-6
-4
-2
Drift (in.)
.
115
The same calculations are repeated for Test 4D (Table 2.1). Figure 2.63 shows
the overall deformed shape of the column at the maximum column drift. Reverse
curvature is noted in the top and bottom hinging regions in Fig. 2.63. The deformations
in the hinging regions are compared in Figs. 2.64 and 2.65 with the associated column
images at the time of maximum column drift. For Test 4D, nonlinear deformation is
noted very clearly in the top hinging region.
116
Bent3 East Column

Maximum Deformed
Shape
(Test4D)
80
Middle Line
Left Line
Right Line
70
60
Column Height (in.)
50
40
30
20
10
0
-6 -4 -2 0
2 4
Drift (in.)
.
117
Middle Line
Top Grid Deformed Shape

(Test4D)
Left Line
Right Line
71
Column Height (in.)
69
67
65
63
61
59
-6
-4
-2
Drift (in.)
.
Bottom Grid Deformed Shape

(Test4D)
Middle Line
Left Line
18
Right Line
16
Column Height (in.)
14
12
10
8
6
4
2
0
-6
-4
-2
Drift (in.)
.
118
The overall deformed shape of the column at the maximum column drift is shown
for Test 6 in Fig. 2.66. Reverse curvature is profoundly noted for Test 6, similar to the
deformed shape for Test 4D. The deformations in the hinging regions are compared in
Figs. 2.67 and 2.68 with the associated column images at the time of maximum column
drift. The deformations represent the images very well.
119
Bent3 East Column

Maximum Deformed
Shape
(Test 6)
Middle Line
Left Line
Right Line
80
70
60
Column Height (in.)
50
40
30
20
10
0
-6 -4 -2 0 2 4 6
Drift (in.)
.
120
Top Grid Deformed Shape

(Test 6)
Middle Line
Left Line
Right Line
73
Column Height (in.)
71
69
67
65
63
61
59
-6
-4
-2
0
Drift (in.)
Bottom Grid Deformed Shape

(Test 6)
Middle Line
Left Line
18
Right Line
Column Height (in.)
16
14
12
10
8
6
4
2
0
-6
-4
-2
0
2
Drift (in.)
121
2.3.4.3 Rotations
As described previously, several methods were evaluated to define the unique
points to calculate the displacements and rotation results from photogrammetry data.
First, using only the even-numbered lines from each frame was used to capture a more
consistent image (without double lines) for analysis. Then, a definition of vertical and
horizontal grid lines was determined. In the first definition, the mid-point of each 0.25in. (6 mm)-thick vertical and horizontal grid line was sought and a middle line was
defined as the connection of these points. In the second definition, the edge lines were
analyzed to obtain the four corner points at each intersecting thick grid line, and also the
average middle point between the corner points. In addition, the length used to define the
edge lines was examined in detail. After all these trials, using the even-numbered lines
from each image, with the average middle of the corner points, with 0.5 in. (13 mm) 1.5
in. (38 mm) in length provided the most consistent results. Using this approach defined
by the edge-lines of the thick grid lines to calculate the rotations of horizontal line at a
given height of the column, there is a large amount of pixelization noted (Fig. 2.51) and
it is difficult to follow the periodicity of the earthquake shakings. By contrast, rotations
of vertical lines provided smoother traces than the rotations of horizontal lines in the
photogrammetry data (Fig. 2.52). It is important to note that the rotations of vertical
lines are average rotations over a given column height defined by the monitored Points.
This section will describe the best final rotation results.
The rotations of vertical lines calculated using Points 3 and 8 and Points 8 and 13
on the bottom grid are shown in Figs. 2.69 and 2.70 for Test 2. There is a small amount
of constant noise (pixelization) in the figures because the amplitude of motion in Test 2
is small. Figures 2.71 and 2.72 show the rotations of vertical lines at the top grid level
122
(Points 38 and 45 and Points 45 and 52) during Test 2. (The general periodicity of
motion can be decerned among the constant noise at low amplitudes of motion.)
0.004
0.003
0.002
Rotation (rad)
0.001
0
0
10
15
20
25
30
35
-0.001
-0.002
-0.003
-0.004
Time (sec.)
Fig. 2.69 Rotation of vertical line calculated using Point 3 and 8 for Test 2
0.01
0.008
0.006
Rotation (rad)
0.004
0.002
0
0
10
15
20
25
30
-0.002
-0.004
-0.006
-0.008
Time (sec.)
123
35
0.008
0.006
0.004
Rotation (rad)
0.002
0
0
10
15
20
25
30
35
-0.002
-0.004
-0.006
-0.008
-0.01
Time (sec.)
Fig. 2.71 - Rotation of vertical line calculated using Point 38 and 45 for Test 2
0.01
0.008
0.006
Rotation (rad)
0.004
0.002
0
0
10
15
20
25
30
-0.002
-0.004
-0.006
-0.008
-0.01
Time (sec.)
Fig. 2.72 - Rotation of vertical line calculated using Point 45 and 52 for Test 2
124
35
The rotations of vertical line calculated from Points 3 and 8 and Point 8 and 13
for Test 4D are shown in Figures 2.73 and 2.74. Figures 2.75 and 2.76 show the
rotations of vertical line at the top grid level for Test 4D.
Figures 2.77 and 2.78 show the results for the bottom hinging region for Test 6,
and Figures 2.79 and 2.80 show for the top hinging region. Both figures provide similar
peak values and periodicity.
0.02
Rotation (rad)
0.015
0.01
0.005
0
0
10
15
20
25
30
-0.005
-0.01
Time (sec.)
Fig. 2.73 - Rotation of vertical line calculated using Point 3 and 8 for Test 4D
125
35
0.03
0.025
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
10
15
20
25
30
35
-0.01
-0.015
-0.02
Time (sec.)
Fig. 2.74 Rotation of vertical line calculated using Point 8 and 13 for Test 4D
0.03
0.025
0.02
0.015
Rotation (rad)
0.01
0.005
0
0
10
15
20
25
30
-0.005
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
Fig. 2.75 Rotation of vertical line calculated using Point 38 and 45 for Test4D
126
35
0.03
0.025
0.02
0.015
Rotation (rad)
0.01
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
Fig. 2.76 Rotation of vertical line calculated using Point 45 and 52 for Test4D
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec.)
127
35
0.06
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec.)
0.08
0.07
0.06
Rotation (rad)
0.05
0.04
0.03
0.02
0.01
0
-0.01 0
10
15
20
25
30
-0.02
-0.03
Time (sec.)
128
35
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec.)
To determine the relative success of the photogrammetry methods to capture

rotations, the maximum rotations of vertical line calculated using photogrammetry
measurements for the three different earthquake motions are normalized with the
maximum cross-sectional rotations obtained from LVDT data. Figure 2.81 shows the
ratios for the three different motions at the two hinging regions with each region having
two different point levels plotted in terms of peak ground acceleration (PGA) for Test2,
Test 4D, and Test 6. As shown in the figure, the ratios vary greatly, with the ratio ranges
in the bottom hinging region between 0.91 in Test 6 and 1.31 in Test 2, and in the top
hinging region ranges between 1.11 in Test 6 and 4.75 in Test 2. The results for the top
and bottom hinging regions well represent the LVDT data in Test 6. The results in the
top hinging region are greatly amplified from the LVDT results for Tests 2 and 4D.
Even with these large differences between the calculated vertical rotation and LVDT data
calculation, it is interesting to note that the maximum rotations represented by vertical
129
rotations using the photogrammetry data closely match the maximum values calculated
from cross-sectional rotations (Fig. 2.55).
5
4.5
5 in.-Bottom
max-PH /max-LVDT
12 in.-Bottom
5 in.-Top
3.5
12 in. Top
3
2.5
2
1.5
1
0.5
0
0
0.5
1.5
PGA (g)
Fig. 2.81 Maximum rotation ratios between photogrammetry compared LVDT results
.
2.4
CONCLUSION
For each deformation calculation, it is necessary to have a consistent and
predictable definition of points on the column surface. The quality of images captured
during the test and the similar pixel intensities across the grid line thickness presented
challenges for calculating consistent displacement and rotation results. The following
conclusions are made regarding the analysis of the deformations in the hinging region
during response to strong ground shaking:
The best image quality is obtained by using the even-numbered lines from
each frame in time.
130
The edges of the 0.25-in. (6 mm) thick grid lines are best defined using a
constant Robert threshold value.
The best definition of a unique point on the column surface is obtained by

using the edge-lines of the 0.25-in. (6 mm) thick grid lines to define four
corner points at the intersection of each horizontal and vertical grid line, and
then averaging these points to define a middle point. The edge-lines are
best defined using lengths of 0.5-1.5 in. (13 mm 38 mm).
The periodicity and amplitude of the displacements calculated at the top of

the column correlate very well with the combined components of DT7, DS5
and DS1 transducer measurements.
Overall, displacement calculations at each grid point follow the periodicity of

motion captured by the displacement transducers very well.
The rotations of vertical line calculated using photogrammetry data, as an

average rotation between two consecutive grid points, represent the crosssectional rotations calculated using the LVDT data in periodicity very well.
The traces are smoother than the detailed rotation calculation of horizontal
lines for photogrammetry data.
The maximum amplitude of rotation is
similar between rotation of vertical line and rotation of horizontal line

calculated using the photogrammetry data.
Suggested changes in future experimental setups to increase the quality of the
images for photogrammetry analysis include:
The edge of the vertical and horizontal grid lines should be much sharper,
with distinct change in intensity from the black to white regions in the
grid.
131
A circular component should be added to the grid region to aid in

determining the rotation of the column. Any changes in the curvature to the
circle would provide additional information on the rotational motion of the
column.
Cameras that operate using progressive scan rather than interlaced images
when recording the events should be used.
132
MANUSCRIPT 1: PHOTOGRAMMETRIC MEASUREMENTS

OF RC BRIDGE COLUMN DEFORMATIONS
a,*
Zeynep Firat Alemdar , JoAnn Browningb and Jeffrey Olafsenc
1147 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
USA
c
Department of Physics, Baylor University, Texas, USA
Corresponding author Tel: +1 785 864 3760 Fax: +1 785 864 5631 E-mail addresses: zeynep@ku.edu (Z.
F. Alemdar).
ABSTRACT:
The determination of the location of nonlinear response in structural systems is
an important step to predict the performance of the system under different loading
conditions. In bridge columns, these nonlinear deformations generally occur over a finite
hinge length. A model of hinging behavior in reinforced concrete bridge columns will
help guide proportioning, detailing and drift estimates for performance-based design.
Data was collected during the NEESR investigation of the seismic performance of fourspan large-scale bridge systems at the University of Nevada Reno that details
deformations in column hinging regions during response to strong shaking events. A
photogrammetry method was applied using a reference grid on the top and bottom
column surfaces to accurately record and analyze deformations in the plastic hinging
regions. The method of application in this study holds several advantages over traditional
sensors including that it is a remote visualization technique, inexpensive and simple to
133
analyze the results. The surface deformations and rotations of a reinforced concrete
bridge column under dynamic loading has been examined and compared with the results
obtained from traditional instruments.
KEYWORDS: Reinforced concrete bridge column, photogrammetry, hinging region,

earthquake loading, frequency domain error.
1. Introduction
This research aims to record and analyze deformations in plastic hinging

locations in concrete bridge structures through an investigation of actual hinging
behavior in large-scale tests of bridge systems subjected to multiple excitations. In order
to evaluate the behavior of plastic hinging regions, a photogrammetric method was used
to remotely track deformations of the concrete surface in the hinging regions.
Photogrammetry is a non-invasive technique of remote visualization of the target
components and a computer rendering of the motion [1]. The tracked motion of the target
components, which were discrete points on the column surface along the joint regions,
was reduced using a computer program to obtain the lateral displacements of the points
and to calculate the vertical and cross-sectional rotation between these points.
Photogrammetry measurements were recorded at the bottom and top hinging regions of
the column. The analysis of simple and inexpensive photogrammetric data was the main
focus of this study.
134
2. Experimental program
Previous research of reinforced concrete column hinging behavior has focused on

small scale component tests under static loadings. Most of these tests were of cantilever
columns in single curvature response [2-15]. Only two studies have included dynamic
tests of reinforced concrete bridge columns to determine the deformations in hinging
regions [16-17].
The four-span reinforced concrete bridge that was tested at the
University of Nevada Reno (UNR) laboratory on February 12-15, 2007 was significant
because it was a large-scale bridge system with columns in double curvature response. In
addition, massive earthquake motions were applied in the longitudinal and transverse
directions of the bridge.
Traditional instrumentation that have been widely used to collect deformation
data from static and dynamic tests include LVDTs (linear variable differential
transformers), displacement transducers, accelerometers, and strain gages. These
instruments are attached to the concrete surface and require electrical connection to a
data acquisition system. As damage occurs to concrete elements, the cover concrete can
spall, gages can fail, and data may be lost. A newer method used to collect deformation
data is by photogrammetry with an advanced non-contact measurement system such as
Krypton system. These systems consist of a suite of cameras, a set of infrared LEDs
(light-emitting diodes), and control software. For data acquisition during testing, the
Krypton cameras are mounted on a reference frame away from the specimen and LEDs
are attached to the specimen surface. The three-dimensional motions of infrared LEDs
(light-emitting diode) are tracked during testing. Several studies have been completed
using the Krypton LED-based 3D displacement measuring system [18-21]. Reducing and
organizing the large amount of data that is collected from this system, hovewer, is
135
difficult. Other limitations of using such advanced systems include the possibility of lost
LEDs during testing due to specimen damage, and the prohibitive cost of the system for
many structural engineering laboratories.
The four-span large scale reinforced concrete bridge test provided a platform to
compare traditional instrumentation measurements with those obtained using a simple
photogrammetry method. The photogrammetry method required only the application of a
reference grid on the concrete surface and a system of cameras to record the motion of
the bridge system. The photogrammetry data showed that the accuracy of the
measurements was approximately 0.04 in. (1 mm). The application of the
photogrammetry method held several advantages over traditional sensors. The cameras
were more economical than those used in other advanced non-contact measuring systems
and could be used in any laboratory. The photographic data did not rely on any physical
connection to the specimen. In addition, any concrete marking that were lost from
excessive deformations could be reconstructed to a certain extent based on surrounding
elements in the frames.
2.1 Specimen details
The test specimen consisted of the four-span reinforced concrete bridge with end
abutments as shown in Fig. 1. The bridge was quarter-scale with two 29-ft (8.84 m)
interior spans, and two 24.5-ft (7.47 m) exterior spans for a total length of approximately
110 ft (33.5 m). The clear heights of the bents were 5, 6 and 7 ft (1.52, 1.83, and 2.13 m),
with the tallest bent in the middle (Bent 2). The superstructure consisted of a solid slab
that was post-tensioned in both the longitudinal and transverse directions. The bridge
136
also included abutment seats at both ends that were driven in the longitudinal direction
by dynamic hydraulic actuators. The depth of the cap beams was 15 in. (0.38 m). The
total length of the cap beams (perpendicular to the longitudinal axis of the bridge) was 98
in. (2.49 m). The concrete compressive strength used in the bridge (excluding the slab)
was 6.7 ksi (46.2 MPa) and reinforcement was ASTM A706 Grade 60. The bents were
numbered Bent 1, Bent 2, and Bent 3 starting from left (South) to right (North).
N
282 in. [7.16 m]
348 in. [8.84 m]
348 in. [8.84 m]
282 in. [7.16 m]

14 in. [0.36 m]
SHAKE TABLE
Abutment 1
SHAKE TABLE
Bent 1
Bent 2
SHAKE TABLE
Bent 3
Abutment 2
Fig. 1 Elevation view of the four-span bridge
2.2 Photogrammetric application setup
Photogrammetry data was collected at the top and bottom hinging regions of the
east column of Bent 3 during the tests. Two grid systems were applied to the surface of
the column as shown in Fig. 2. The bottom grid consisted of vertical and horizontal lines
spaced irregularly around the face of the circular column. Deformation data was
collected at various points of the intersecting lines. The top grid system used a
combination of lines and squares, having the same spacing as the lines in the bottom
grid. This top grid pattern was applied to simulate targets on the column surface.
137
To apply the grid systems, the column was first spray-painted black for a solid
dark background. The lines and squares were lightly traced with a pencil, and then a
single stencil was used to fill in the lines and squares using white paint.
(a)
(
b)
Fig. 2 Close view of (a) bottom and (b) top grid systems
A series of linear vertical displacement transducers (LVDTs) were used to record

deformations in the transverse and longitudinal direction of the bridge, as shown in Fig.
2. One displacement transducer (DT7) was located at the middle of the t-beam on the top
of the cap beam (Fig. 3). The LVDT and DT7 data were collected by the research team at
UNR.
138
DT7
3ETR3
57
56 58
50
49 51
59
52
62
6061
55
5354
43
42 44
45
48
4647
3ETR5
3ETR6
36
35 37
29
28 30
212223
41
38 3940
34
31 3233
24 252627
1617 18
19 20
1112
13
1415
6 7
9 10
1 2
4 5
3EBR5
3EBR7
3ETR4
3EBR6
3EBR8
Fig. 3 Grid system and LVDT locations on column in the Bent 3 east column
An aluminum tower was built to hold the cameras that would record the column
deformations caused by the earthquake events (Fig. 4). The tower had two levels to allow
recordings at the top and bottom plastic hinging regions. A system of two cameras
(DXB-9212EF model Starlight 600 TVL super high resolution) with black-white
recording properties was mounted on each level [at 94.5 in. (2400 mm) and 159 in. (4039
mm) height]. Video-lens zooms had 60-300 mm F4-5.6 capturing capability.
139
Fig. 4 The aluminum tower and four cameras
2.3 Earthquake loading
The motions that were used for the shaking table tests were based on the 1994
Northridge earthquake as measured at the ground station at the Century City Country
Club North. The Century City station is owned by the California Strong Motion
Instrumentation Program, and is located at 34.063 latitude, -118.418 longitude. Both the
90 degree and 360 degree lateral components were used in the tests. The same motion
was used for each test, but the peak ground acceleration (PGA) of the motion was
amplified as additional tests were completed. Over the course of the tests, the PGA of the
original motion was scaled from 0.075g to 1.0g in transverse direction (90 degree
component) and from 0.09g to 1.2g in longitudinal direction (360 degree component) as
140
shown in Table 1. The earthquake motion was applied for uniaxial and biaxial tests of
the structural system. The motion was applied 13 times with six different scaled
intensities so that the progression of damage would be evaluated from pre-yield to failure
during the shaking tests.
Table 1 Earthquake motion levels applied during the tests
Test No.
Test Date
1A
Motion
Level
Test Type
Motion PGA (g)

Trans.
Long.
Longitudinal
0.09
Longitudinal
0.09
1C
Longitudinal
0.09
1D
Biaxial
0.075
0.09
Biaxial
0.15
0.18
Biaxial
Longitudinal
0.6
4B
Longitudinal
0.6
4C
Longitudinal
0.6
4D
Biaxial
0.5
0.6
Biaxial
0.75
0.9
Biaxial
1.2
Biaxial
1.2
1B
4A
5
6
6
2/12/2007
2/13/2007
2/14/2007
2/15/2007
3. Definition of unique points on column surface
The primary focus of this investigation was to determine displacements and

rotations of points on the column surface during the earthquake loading. This was only
possible by having a consistent and predictable definition of points, which was
complicated by several degraded images and ambiguous locations of precise points on
the painted surface. In addition, there were some inherent differences in the data obtained
with LVDT instrumentation and that recorded for the photogrammetry application. One
of these differences was that the rotation calculations obtained using photogrammetry
141
analysis represented the rotations of horizontal lines based on the movement of the
column, which included flexural and rigid body components. Rotations calculated using
the LVDT measurements from the test included only flexural contributions. Another
difference was that the LVDT instrumentation was set to record movement in the
transverse and longitudinal planes, whereas the photogrammetry measurements were
recorded at a horizontal angle between the orthogonal directions, as selected with the
setup of the aluminum tower and the cameras (Fig. 4). These differences are discussed
and accounted for in the analysis and results.
A measure of the best definition of unique points and deformations was made
using a FDE (frequency domain error) index as developed by Dragovich and Lepage
[22]. The FDE index is used to compare time-domain response waveforms such as those
described in this study. The FDE method uses the Fourier spectra of the response signals,
which are non-periodic and comprised of multiple frequencies, to calculate the index.
Fast Fourier Transform was used to determine the composition of frequencies of the
photogrammetry responses. The starting and ending frequencies were selected to be
between 0.5 Hz and the Nyquist frequency, which is fn =1/(2t) with t as the time
interval of the series. The FDE index accounts for both the amplitude and phase
differences between the response signals and gives a number between 0 and 1 as a
measure of the goodness-of-fit, where zero indicates a perfect correlation. The amplitude
error is a measure of the difference between the values of the response signals, and the
phase angle is the angle that the resultant vector of the signals makes with respect to the
real axis in the real-complex plane. From the study by Dragovich and Lepage, an FDE
value equal to 0.75 represented a poor correlation, and an FDE value of 0.25 represented
a very good correlation. These correlations, however, were based on analytical responses
142
of reinforced concrete frames, and comparisons of two experimental measurements

would be expected to have more noise than analytical responses.
3.1. Challenges with unclear images
Interactive Data Language (IDL) was used to analyze the video images that were
recorded during the test. Each video was first divided into over one thousand images for
individual analysis. The displacement of a unique point in an image was defined as the
relative horizontal movement of that point between successive images. At times,
however, it was difficult to keep track of a unique point because of the quality of the
images.
Each image (640 x 480) was written as 480 lines that were 640 pixels wide, and
represented 1/30th of a second during the test. The images were interlaced: the oddnumbered lines are written first, followed by the even-numbered lines so that the image
was really two images offset by 1/60th of a second. The first partial image (half of the
total image) was comprised of the odd lines, and the second partial image was
comprised of the even lines. Due to shaking of the strong floor during the tests and the
incompatibility in recording frequency between the cameras and the earthquake motions,
several images were degraded in quality (corrupted). As a result, some of the horizontal
and vertical lines appeared as double lines, and some images were simply blurry. In
future tests, cameras that operate with progressive scan rather than interlaced images
should be employed to avoid these problems.
The even-numbered and odd-numbered lines of the images were considered
separately to improve the calculations of displacements and rotations with the corrupted
images. When the image was divided into separate images of even and odd lines, the
143
double image effect that was produced by a shift between the two halves of the image
being written was eliminated. The vertical displacement results of Point 7 (located in Fig.
3) from the even and odd line analyses were reduced from the Test 4D motion (Table 1).
A comparison of the calculated response obtained from the photogrammetry
analysis and the measured response from traditional instruments was completed. The
calculated photogrammetric response was sensitive to how unique points on the column
surface were defined. The quality of the images during the shaking events affected the
final representation of the column deformations. The vertical displacement results of
Point 7 are compared with the LVDT 3EBR7 readings in Figs. 4 and 5 for even and odd
line analyses, respectively. FDE indexes were calculated for the vertical displacements of
Point 7 in the even and odd analysis as compared to LVDT 3EBR7 results. Table 2
shows that the amplitude error is very good at only 0.16 for the even line analysis,
whereas it is 0.19 for the odd line analysis. The phase errors (periodicity errors) are very
similar at around 0.50. Overall, the total FDE index is slightly smaller for the even line
analysis than the odd lines, and is indicative of which set of lines (even/odd) captured
first.
Table 2 FDE index results for Point 7
FDE Results
Comparisons
Fig. 7-Even line
analysis vs. LVDT
Fig. 8-Odd line
analysis vs. LVDT
Amplitude
Phase
Total
0.16
0.51
0.67
0.19
0.49
0.68
0.12
0.58
0.70
0.10
0.45
0.55
Fig. 11-Vertical vs.

horizontal line
rotation
Fig. 12-Photo vs.
Displacement
144
Fig. 5 Comparison of Point 7 vertical displacement for even lines with LVDT 3EBR7
data
Fig. 6 Comparison of Point 7 vertical displacement for odd lines with LVDT 3EBR7 data
145
3.2. Intersecting of vertical and horizontal grid lines
The definition of each unique point on the grid surface was a critical step to
obtain the displacements of the points and rotations of sections of the column. The
intersection of the thick vertical and horizontal lines (0.25 in. thick) was used to define
the points, however, there were several ways to consider creating a line within that
thickness. One option was to find the mid-point of each 0.25-in. thickness and define a
middle line with the connection of those points (Fig. 6). The pixel that had the
maximum intensity value in the intersections of the horizontal and vertical lines was
defined as the mid-point within the thickness.
The edges of the thick grid lines were considered as a second option to define two
new lines (edge-lines) for each horizontal and vertical grid line. Using this option, there
were four intersections of the edge lines to consider as possible unique locations of that
Point, as seen in Fig. 7. In this method, the edge lines may be determined by points of
a particular light intensity value, a Robert Threshold, or as a considerable difference in
relative intensity between the black to white paint transition. The length of the line also
had an effect on the results. Longer [1.5 in. (38 mm) to 3.0 in. (76 mm)] and shorter [0.5
in. (13 mm) to 1.5 in. (38 mm)] edge-line lengths were investigated to determine the
influence on the definition of a unique Point.
The best representation of a unique Point was obtained from the average of four
corner point coordinates. First, the four corner points at the intersection of each
horizontal and vertical grid line were defined using the edge-lines of the 0.25-in.-thick
grid lines. These edge lines were defined using a constant Robert threshold value of 50
with a short edge line length that ranged from 0.5-1.5 in. The coordinates of the four
corner points were then averaged to define a middle point.
146
Middle Line
Edge Lines
Edge Point
Middle Point
Fig. 7 Close up of lines used to define Point 3 (Fig. 3)
0.25 in. [0.01 m]

a
0.25 in. [0.01 m]
0.25 in. [0.01 m]

a
0.25 in. [0.01 m]
13
12
c
~2.5 in. [0.06 m]
Fig. 8 Four corners surrounding general location of Point 12 and 13
3.3. Comparison of rotation of vertical line and rotation of horizontal line
It is of interest to compare the rotations of lines defined by the relative movement

of two points that are aligned vertically (representing an average cross sectional rotation
over the height bounded by the two points) and two points that are aligned horizontally
(representing the cross sectional rotation at the height of the two points). The rotation of
a horizontal line on the column bottom surface was calculated using the even-numbered
lines from the photogrammetry measurements. The rotation of a horizontal line was
approximated as the relative vertical displacement between two points divided by the
horizontal distance between these points. The change in horizontal distance between the
147
points within each image was accounted for in the analysis. The rotation of a vertical line
on the column was approximated using the relative transverse displacements of two
points as reduced from photogrammetry data that were aligned vertically on the grid
surface divided by the length between those lines. This rotation represents the average
cross-sectional rotation between those two points if shear deformations can be neglected.
Figure 8 shows the rotation of a vertical line that was calculated using the lateral
displacements of Points 3 and 8 on the bottom grid as were tracked during the Test 4D
motion. Figure 9 shows the rotation of a horizontal line located at the top of the region
encompassed by the vertical line (calculated using the vertical displacements of Points 7
and 9).
The comparison between the rotations of the vertical line shown in Fig. 8 and the
horizontal line shown in Fig. 9 is shown in Fig. 10. Although the top and bottom hinging
regions had numerous cracks and some spalling of the concrete on the column surface, to
the extent that a few lateral reinforcing bars and yielding of the longitudinal bars were
revealed, the photogrammetry method was able to capture the rotations well. As an
average measurement, the rotation of the vertical line provides a smoother trace than the
rotation of the horizontal line that was calculated with the photogrammetry data. The
correlation between the rotation of vertical and horizontal line was examined using the
FDE index (Table 2), and the amplitude error was found to be 0.12, which represents a
very good correlation. The phase error, however, was 0.58 and the total FDE index was
0.70, which overall was a fair correlation.
148
0.03
0.025
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
-0.01
-0.015
-0.02
0
10
15
20
Time (sec)
25
30
35
40
35
40
Fig. 9 Rotation of a vertical line calculated using Point 3 and 8
0.03
0.025
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
-0.01
-0.015
-0.02
0
10
15
20
25
30
Time (sec)
Fig. 10 Rotation of a horizontal line using Points 7 and 9
149
0.03
Rotation of a Horizontal Line
0.025
Rotation of a Vertical Line
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
Rotation (rad)
-0.01
-0.015
-0.02
0
10
15
20
Time (sec)
0.03
0.02
0.01
0
-0.01
-0.02
25
30
8
10
35
T ime (sec)
Fig. 11 Comparison of rotations (local vs. average) in Test 4D
3.4. Determining the deformed shape of the column hinges
The grid system used to collect photogrammetry data is advantageous when

determining plastic hinge lengths as compared to traditional measurements garnered
from strain gages and LVDTs. With photogrammetry, the deformation of the column
surface can be determined over a continuum, whereas LVDT instrumentation provides
displacement data for discrete points. In addition, if significant damage occurs to the
concrete surface LVDT measurements may become dysfunctional, but the painted grid
surface may be reconstructed in an image and photogrammetry data preserved during the
tests.
The transverse and vertical displacements of each grid point along the top and
bottom hinging surfaces were calculated for each test (Table 1). Point 59 in the top grid
150
system (Fig. 3), which was the closest point to the top of the column, was selected to
track the transverse displacement along the direction of the focal plane of the cameras on
the left side of the aluminum tower. The projection of the camera view from the tower
had an angle of 15.25 with the longitudinal direction of the bridge (Fig. 4). The
transverse displacements of grid Point 59 captured during Test 4D was compared with
the combined components of the displacement transducers mounted on the bridge deck
and the north abutment as shown in Fig. 11 and described next. First, the measurements
recorded by DT7 (transverse displacement of Bent 3 at mid-height of the bridge deck)
were used to obtain the transverse component along the focal plane of the cameras. Then,
DS5 and DS1 measurements were combined together to find the longitudinal component
of the displacement (DS5 recorded the gap size between the north abutment and the
bridge deck while DS1 measured the movements of the north abutment in the
longitudinal direction of the bridge). Although there was approximately 23.3 in. (592
mm) between the transverse beam and Point 59, the column deformed in double
curvature and the lateral displacements at these two heights were assumed to be similar.
As seen in Fig. 11, the overall periodicity of the two measurements correlates very well,
with differences in magnitude primarily attributed to small flexibilities of the transverse
beam creating differential movement between the top beam and Point 59 in the column
hinging region.
The FDE index analysis was used to compare the results of transverse
displacements at Point 59 and the movement recorded by the displacement transducers.
The amplitude error was found to be 0.10, which represented a very good correlation
between the photogrammetry and displacement transducers results (Table 2). This is
significant as the maximum displacement of bridge columns is an important parameter
151
used to qualify expected damage during earthquake events. The phase angle error was
calculated to be 0.45, which was a good correlation result.
The deformed shapes of the hinging regions were defined over a continuum by
considering the movements of a suite of Points in the regions [23]. The displacement
calculations were repeated at each of the grid points along the hinging surface. Three
vertical lines, as defined by the grids shown in Fig. 3, were monitored in Test 4D. On the
middle line, the selected Points were 3, 8, 13, and 18 for the bottom grid and 38, 45, 52,
and 59 for the top grid system. For the left line, Points 2, 7, 12, and 17 were chosen at
the bottom and Points 37, 44, 51 and 58 were selected at the top. For the right line,
displacements of Top Points 4, 9, 14, and 19 and Bottom Points 39, 46, 53, and 60 were
monitored. The deformations in the top and bottom hinging regions during Test 4D, as
represented by dotted lines connecting the displaced points, are compared in Figs. 12
and 13 with the associated column images at the time of maximum column drift. Reverse
curvature is correctly represented in the top and bottom hinging regions. Slight nonlinear
deformation is also visible in the top hinging region. The deformations captured by the
photogrammetry data represent the images very well as seen in Figs. 12 and 13.
152
80
Photogrammetry
components
60
Displacement (mm)
40
20
0
0
10
15
20
25
30
35
40
-20
-40
Time (sec.)
Fig. 12 Comparison of Drift at Point 59 for Test 4D
1850
1800
Column Height (mm)
1750
1700
1650
1600
1550
1500
-50
50
100
150
Horizontal Position (mm)
Fig. 13 Top grid deformed shape with picture comparison at maximum column drift
during Test 4D
153
450
400
Column Height (mm)
350
300
250
200
150
100
50
-100
-50
50
100
150
Horizontal Position (mm)
Fig. 14 Bottom grid deformed shape with picture comparison at maximum column drift
during Test 4D
4. Conclusions
A simple photogrammetry analysis method was applied to evaluate the

deformations of a bridge column surface in plastic hinging locations during the response
to strong ground motion. The definition of points on the column surface was completed
by using the edge-lines of the 0.25-in. (6 mm) thick grid lines to define four corner
points at the intersection of each horizontal and vertical grid line, and then averaging
these points to define a middle point. The edge-lines were best defined using a
constant Robert threshold value and short line lengths of 0.5-1.5 in. (13 mm 38 mm).
The even-numbered lines from each frame (partial image) were selected to improve the
quality of the images recorded during the tests. The following conclusions are made
regarding the analysis of the deformations in the hinging region during response to
strong ground shaking:
154
The FDE indexes obtained for the vertical displacements of Point 7 in the
even and odd analysis gave very good correlations between the calculated
amplitude using photogrammetry data and the measured responses from the
LVDT 3EBR7 recordings. The phase error results had fair FDE values
(approximately 0.50), and fair overall values (0.67 and 0.68).
The rotation of a vertical line calculated using photogrammetry data, as an

average representation of cross-sectional rotation, provided a better
representation of the column rotation than the rotation defined by a horizontal
line. The correlation between the amplitudes of the rotations of horizontal and
vertical lines gave very good results as found in the FDE index calculations.
The overall correlation was fair because of differences in phase angles.
The amplitude of the displacements calculated at the top of the column

correlate very well with the results from the combinations of DT7, DS5 and
DS1 transducer measurements at the superstructure of the bridge as
represented in the FDE index calculation. The overall correlation was good.
The deformed shape of the top and bottom hinging regions was constructed
using the displacement of the points on the column surface. The deformed
shapes show a good match with the images at the maximum drift of Test 4D
by correctly representing double curvature of the column and the general
deformed shape of the hinging region.
155
Acknowledgements
This project gratefully acknowledges the support of NSF grant #0532084, Joy
Pauschke Program Director. The assistance and support of Prof. Saiid Saiidi, Graduate
Research Assistant Roby Nelson, the entire staff of the research facilities at the
University of Nevada Reno, and Graduate Student Nick Hunt at University of Kansas are
also gratefully acknowledged. Prof. Olafsen was supported in part by a Big XII Faculty
Fellowship. Additional support from KU Transportation Research Institute is gratefully
acknowledged.
References
[1] Jauregui D., Ruinian J., K. White. Review of close-range photogrammetry

applications in bridge engineering. Washington D. C. Transportation Research
Board, 2006.
[2] Baker A.L.L. Ultimate load theory applied to the design of reinforced and
prestressed concrete frames. Concrete Publications Ltd., London. 1956.
[3] Baker A.L.L., Amarakone A.M.N. Inelastic hyperstatic frame analysis. Proc.
International Symposium on the Flexural Mechanics of Reinforced Concrete, ACI
SP-12 Miami, pp. 85-142. 1964.
[4] Mattock A.H. Rotational capacity of hinging regions in reinforced concrete
beams. Proc. International Symposium on the Flexural Mechanics of Reinforced
Concrete, ACI SP-12 Miami, pp. 143-181. 1964.
156
[5] Corley W. G. Rotational capacity of reinforced concrete beams. J Struct Division

ASCE 1966; V. 92, No. ST5, pp. 121-146.
[6] Park R., Priestley M. J. N., Gill W. D. Ductility of square-confined concrete
columns. J Struct ASCE 1982; 108 No. ST4. 929-950.
[7] Priestley M. J. N., Park R. Strength and ductility of concrete bridge columns
under seismic loading. ACI Struct J 1987 V.84 Issue 1. pp. 61-76.
[8] Tanaka H., Park R. Effect of lateral confining reinforcement on the ductile
behavior of reinforced concrete columns. Research Report 90-2. Department of Civil
Engineering, University of Canterbury, Christchurch, New Zealand. 458 pp. 1990.
[9] Paulay T., Priestley M. J. N. Seismic design of reinforced concrete and masonry
structures. John Wiley & Sons, Inc. 1992.
[10] Sheikh S.A., Khoury S.S. Confined concrete columns with stubs. ACI Struct J
1993 V. 90 No. 4 pp. 414-431.
[11] Watson S., Park R. Simulated seismic load tests on reinforced concrete columns.
J Struct Engineering, ASCE 1994; V. 120, No. 6, June, pp. 1825-1849.
[12] Kovacic D. Design of high-strength concrete columns. M.Eng. thesis. the
University of Melbourne; 1995.
[13] Mendis P. Plastic hinge lengths of normal and high-strength concrete in flexure.
Advances in Structural Engineering 2001 V. 4 No. 4 pp. 189-195.
[14] Bae S. Seismic performance of full-scale reinforced concrete columns. Ph.D.
dissertation, Austin (TX): the University of Texas; 2005.
[15] Phan V., Saiidi M.S., Anderson J., Ghasemi H. Near-fault ground motion effects
on reinforced concrete bridge columns. Journal of Structural Eng. 2007; V.133, Issue
7, pp. 982-989.
157
[16] Dodd L.L. The dynamic behavior of reinforced concrete bridge piers subjected
to New Zealand seismicity. Ph.D. thesis, the University of Canterbury Christchurch
New Zealand; 1992.
[17] Hachem M. M., Mahin S.A., Moehle J.P. Performance of circular reinforced
concrete bridge columns under bidirectional earthquake loading. Report No. PEER
2003/06, Pac. Earthq. Eng. Research Center, University of California at Berkeley,
Feb. 2003, 490 pp.
[18] Brueggen B., Waugh J., Aaleti S., Johnson B., French C., Sritharan S. et
al. Tests of structural walls to determine deformation contributions of interest for
performance-based design. Proc. of 2007 ASCE Structures Congress, Long Beach,
CA, 2007.
[19] Matamoros A. B., Matchulat L., Woods C. Axial load failure of shear critical
columns subjected to high levels of axial load. Proc. 14th World Conference on
Earthquake Engineering. 2008.
[20] Matchulat Lisa. Mitigation of collapse risk in vulnerable concrete buildings,
M.E., University of Kansas; 2009.
[21] Leon R.T., Kim D.K., Hajjar, J.F. Limit state response of composite columns
and beam-columns. Part 1: Formulation of design provisions for the 2005 AISC
Specification, AISC Engineering Journal, Vol. 44, No. 4, pp. 341-358. 2007.
[22] Dragovich JJ., Lepage A. FDE index for goodness-of-fit between measured and
calculated response signals. Earthquake Engineering and Structural Dynamics
2009;38:1751-1758.
[23] Firat Alemdar Z. Evaluation of plastic hinge regions in reinforced concrete
bridge systems. Ph.D. dissertation, Lawrence (KS): the University of Kansas; 2010.
158
MANUSCRIPT 2: MODELING SURFACE DEFORMATIONS

AND HINGING REGIONS IN REINFORCED CONCRETE
BRIDGE COLUMNS
a,*
Zeynep Firat Alemdar , Adolfo Matamorosb and JoAnn Browningc
USA
USA
USA
Corresponding author Tel: +1 785 864 3760 Fax: +1 785 864 5631 E-mail addresses: zeynep@ku.edu (Z.
F. Alemdar).
1. Introduction
The determination of hinging regions and damage locations in reinforced

concrete bridge columns is a critical step to evaluate the performance of a bridge system
under earthquake events. Bridge systems are designed to keep inelastic behavior in the
columns and away from the superstructure, which is different than building systems. In
bridge systems, the columns of the bridges must generally provide the inelastic rotational
capacity for the structure.
The length of the spread of plastic behavior in columns, or plastic hinge length,
has been studied extensively by many researchers [1-10]. This study investigates the
plastic hinge length in reinforced concrete bridge columns, focusing on a bridge system
159
under dynamic loads. A consistent prediction of a plastic hinge length is necessary to

examine the theoretical drift capacity of bridge columns. The hinge length also indicates
the theoretical length of damage concentration along the column. Previous work has
relied on determining expressions to estimate hinge lengths from small-scale component
tests. An evaluation of large-scale system behavior has the benefit of including the
effects of moment redistribution and progression of yielding throughout the entire
structure. As moments are redistributed, the order of hinges in the bridge system and the
magnitude of plastic deformations in each column can change.
Modeling the magnitude of surface deformations in hinging regions is crucial to
detail the spread of plasticity in the hinging regions during the response of earthquake
loading. In the reported study, a photogrammetry method was used to remotely track
deformations of a concrete bridge column surface in a distributed region rather than at a
specific location. The bridge column was tested as part of a large-scale bridge test at the
University of Nevada Reno (UNR) [11]. The photogrammetry method holds several
advantages over traditional sensors, including that the experimental components are
simple to construct, are more economical for typical laboratory environments, and are
likely to fail less at large deformation than traditional strain measurement devices. In
addition, extensive surface deformations and rotations in the hinging regions can be
collected with this method.
This research uses experimental data from photogrammetry measurements of
bridge column deformations tested at UNR to create a finite element model that
realistically represents deformations in a reinforced concrete bridge pier. Experimental
data was obtained from the application of the photogrammetry method. Cameras were
used to record the movements during earthquake shakings of two grid systems applied to
160
the column. The displacements were reduced using the pictures recorded by the cameras.
The bridge column was analyzed using Finite element (FE) analysis with the ABAQUS
[12] software. Analyses of the bridge column under monotonic and earthquake-induced
deformations were completed, and the load deflection curve under monotonic loading
was compared with a similar curve produced using the Modified Compression Field
Theory. The FE model results under earthquake-induced deflections were compared with
the photogrammetry measurements.
2. Description of four-span large scale reinforced concrete bridge test
A large-scale reinforced concrete bridge system was tested on February 12-15,

2007 at the UNR laboratory under biaxial earthquake loadings. The test specimen
consisted of a four-span reinforced concrete bridge with end abutments as shown in Fig.
1. The bridge was quarter-scale with two interior and two exterior spans that were 29 ft
(8840 mm) and 24.5 ft (7470 mm) in length, respectively. The total length of the bridge
system was approximately 110 ft (33528 mm). The clear heights of the bents were 5, 6
and 7 ft (1524, 1830, and 2130 mm), with the tallest bent in the middle (Bent 2). The
concrete compressive strength used in the bridge (excluding the post-tensioned slab) was
6.7 ksi (46.2 MPa) and the yield strength of the reinforcement was Grade 60 ASTM
A706 (414 MPa) deformed bar. The columns had double curvature response and resisted
high intensity earthquake motions during testing. The superstructure consisted of a posttensioned solid slab both in the longitudinal and transverse directions. The bridge also
had abutment seats at both ends that were driven in the longitudinal direction by dynamic
hydraulic actuators to simulate the gap closure and impact at the abutments.
161
The top and bottom hinging regions of the east column of Bent 3 were the main
targets to collect photogrammetric data in the large-scale four-span reinforced concrete
bridge test. The column that was selected for application of the photogrammetry method
was 72 in. (1829 mm) long and 12 in. (305 mm) in diameter, with a 0.5 in. (13 mm)
concrete cover. The column consisted of 16 No.3 (16-10M) longitudinal reinforcement
equally spaced along the radial direction. The cap beam is 15 in. (381 mm) by 15 in. with
a 23- in. (584 m) width. The total length of the cap beams along the transverse direction
of the bridge was 98 in. (2490 m).
Two different grid systems were applied to track the displacements of the
columns (Fig. 2). To construct these grid systems, the bridge column was first spraypainted black to obtain a solid dark background and the lines and squares were lightly
traced with a pencil, and then they were filled in using a single stencil and white paint.
The bottom grid had vertical and horizontal lines spaced irregularly around the face of
the circular column while the top grid used a different configuration that was composed
of lines and squares so as to be similar to the spacing of the lines in the bottom grid. The
intersections of the thick vertical and horizontal lines in both regions were numbered and
used to obtain the best definition of Points along the column surface after the analysis of
the photogrammetry data as shown in Fig. 3 [13].
The biaxial and uniaxial scaled motions measured at the Century City Country
Club during the 1994 Northridge, California earthquake were applied during the bridge
tests. This earthquake motion was used 13 times with increasing amplitude in order to
monitor the progression of damage from pre-yield to failure. A white noise excitation
was also applied before each ground motion intensity change to test for system
162
identification. The un-scaled lateral and longitudinal components of Northridge

earthquake motion were given in Figs. 4 and 5.
An aluminum tower was built to place cameras to record the earthquake events
during the shaking tests. The tower had two levels to monitor of the two plastic hinging
regions. DXB-9212EF model Starlight 600 TVL super high resolution cameras with
black-white recording properties were placed on each level of the tower and video-lens
zooms having 60-300mm F4-5.6 capturing capability were used with the cameras to
clearly see the grid surfaces as shown in Fig. 6. Two monitors (two channels for each
monitor) were set up to show the field of view of the cameras during the tests. The
distance between the east column of Bent 3 and the tower was 288 in (7315 mm). The
angles from the left and right cameras to the column were 73.58 and 44.97 on the
horizontal plane as shown in Fig. 7, respectively. Therefore, the displacements from
photogrammetric measurements were obtained along the direction of the focal plane and
compared with the combined displacements of transverse and longitudinal displacement
transducers on the superstructure of the bridge.
3. Finite element model of the bridge column
A three dimensional finite element model of the east column of Bent 3 was
defined with the cap beam on the top of the column and the footing system under the
circular column using ABAQUS Finite Element software. A concrete damaged plasticity
model was utilized in the FE analysis to define the uniaxial compressive and tensile
concrete material properties of the bridge column, while elastic concrete material
properties were assumed for the cap beam and footing of the system. The properties of
163
the concrete in the core and cover were defined using a plasticity based material model.
The stress-strain curve in compression was defined using the Mander unified stress-strain
model under monotonic loading at slow strain rates [14] as shown in Fig. 8. The
maximum strain was selected to be 0.1, similar to the research done by Scott et al.
(1982), Spacone et al. (1996), and Yassin (1994) who used the maximum concrete strain
in compression as 0.06. The response of concrete subjected to the moderately rapid load
rates, as associated with earthquake ground motions, was also considered. The dynamic
magnification factors that were used in the analysis followed the recommendations of
Mander (1984), who used regression analyses of the experimental results of Watstein
(1953) on plain concrete specimens of different strengths. The relationships for the
dynamic strength of concrete (Df) are as follows.
(fco)dyn = Df fco
(1)
where:
(fco)dyn = the compressive strength of concrete under rapid loading
fco = the quasi-static compressive strength of concrete in MPa (1 Mpa=145 psi).
1/ 6
c
1+
0.035( f co' ) 2
Df =
1/ 6
0.00001
1+
' 2
0.035( f co )
(2)
where c= rate of strain in s-1.

Dynamic strain rates were also considered for the proposed model of the
reinforced concrete bridge column. Previous work in numerical modeling [19] and
experimental dynamic testing of full-scaled models [20] investigated the strain rate
associated with earthquake loading of reinforced concrete structures. This research and
the investigations of others [21] identified the maximum strain rate in reinforced
164
concrete structures subjected to severe earthquake ground motion to be between 0.001

strain per second and 0.25 strain per second. An appropriate maximum strain-rate for the
case of crushing of plain concrete is estimated assuming that crushing occurs once some
inelastic deformation has occurred in the bridge. A maximum strain rate for
consideration of the concrete crushing strength was assumed to be 0.05 strain per second
according to these investigations.
Stiffness relationships in the FE model were complex. The relationships for the
dynamic stiffness factor (DE) that were used in the model were very similar to the
dynamic strength equations, as shown below:
(Ec)dyn = DE Ec
(3)
1/ 6
c
1+
' 3
0.035( f co )
DE =
1/ 6
0.00001
1+
' 3
0.035( f co )
(4)
where (Ec)dyn = Elasticity modulus of concrete under dynamic loading

c = Compressive strain of concrete (in./in.).
The stress-strain curve for concrete in uniaxial tension was studied using the
equations proposed by Bhide (1987), Vecchio and Collins (1982), and Collins and
Mitchell (1987). These three different tension models are compared in Fig. 9. The Bhide
(1987) tension model was chosen for this study because it provided the best
convergence. The model also takes into account the ability of the concrete to have a
tension stiffening effect. To model this behavior, the modulus of rupture was taken to be
4.0*(fc). The angle of cracking was assumed to be 35 to obtain smooth decreases in
the tensile stresses after cracking.
165
A uniaxial bilinear steel model with isotropic kinematic hardening properties was
used to simulate the behavior of the longitudinal steel. Isotropic parameters were
implemented using the classical stress-strain relationship for ductile steel according to
ASTM A706 Grade 60 Steel. In order to adjust the steel strength for loading in the range
of strain rates developed under earthquake shakings, dynamic magnification factors were
calculated. The data for A615 steel [21] showed that the dynamic yield strength was
approximately 10 percent larger than the static yield strength. According to
investigations done by Manjoine (1944), the maximum increase in ultimate strength of
mild structural steel was approximately 4%. Several studies [21, 25, and 26] indicated
that increased strain rate had relatively little effect on the elastic modulus of steel. The
dynamic factor of 10 percent for the yield strength was used in the material model and
compared as shown in Fig. 10.
Kinematic hardening properties were defined including cyclic strain softening as
defined by Ma et al. (1976). A reinforcing steel element was modeled with the cyclic
material properties that had been proposed by Ma et al. (1976) and subjected to reversed
cyclic loading. The stress-strain curve of an element along the steel member was
obtained from the FE analysis and compared with the experimental data conducted by
Ma et al. (1976) as shown in Fig. 11. The FE model represented the results reported by
Ma et al. (1976) very well.
Mesh sensitivity analysis was studied under monotonic loading for the FE model
of the bridge column. The axial load on the column consisted of an imposed weight of
47.2 kips applied as a uniform pressure at the surface of the cap beam in the FE model.
The two-directional shaking table displacement measurements that were recorded by the
research team at UNR were applied at the bottom surface of the footing as transverse and
166
longitudinal direction displacement histories. The circular column, cap beam and the
footing were modeled using 3D continuum 8-node brick elements, whereas longitudinal
reinforcement and transverse steel hoops were defined using 1D 2-node linear truss
elements. The longitudinal steel had a cross-sectional area of 0.11 in.2 (71 mm2) and
transverse hoops had 0.029 in.2 (18.7 mm2) cross-sectional area evenly distributed every
1.25 in. (32 mm) along the column. The brick elements had twenty integration points
distributed throughout its physical volume. A reduced integration scheme was utilized to
reduce the computation time for the analysis.
The FE mesh of the solid elements was generated considering the location of the
longitudinal reinforcement. Truss elements were linked to the edge of the solid elements
using an embedded constraint option. To determine the mesh sensitivity in the FE model,
three different models having mesh sizes of 3 in. (76 mm), 1 in. (25 mm) and 0.5 in. (13
mm) along the top and bottom hinging regions of the column and 4 in. (102 mm) outside
the regions were studied. A coarser mesh was used for the cap beam and the footing
because they were modeled with elastic material properties. The FE model was analyzed
under increasing lateral loads. Figure 12 shows that the 1 in. (25 mm) mesh model gave a
better force-displacement distribution when compared to similar curves generated using
the Modified Compression Field Theory [28] and the result of a moment-curvature
program (Curve). The analysis of the 0.5 in. (13 mm) mesh model along the hinging
regions took 2 times longer than using 1 in. (25 mm) mesh without any benefit in model
performance. The 1 in. mesh size was chosen for future analyses.
The success of the model is measured by how accurately the response of the
structure is represented. To accomplish this measurement, a quantitative procedure is
used to compare measured and calculated response values. This measure was made using
167
a FDE (frequency domain error) index as developed by Dragovich and Lepage [29]. The
FDE index is used to compare time-domain response waveforms such as earthquake
responses obtained in this study. The FDE method uses the Fourier spectra of the
response signals, which are non-periodic and comprised of multiple frequencies, to
calculate the index. Fast Fourier Transform was used to determine the composition of
frequencies of the photogrammetry responses. The starting and ending frequencies were
selected to be between 0.5 Hz and the Nyquist frequency, which is fn =1/(2t) with t as
the time interval of the series. The FDE index accounts for both the amplitude and phase
differences between the response signals and gives a number between 0 and 1 as a
measure of the goodness-of-fit, where zero indicates a perfect correlation. The amplitude
error is a measure of the difference between the values of the response signals, and the
phase angle is the angle that the resultant vector of the signals make with respect to the
real axis in the real-complex plane. From the study by Dragovich and Lepage, an FDE
value equal to 0.75 represented a poor correlation, and an FDE value of 0.25 represented
a very good correlation.
The results of the FE model of the bridge column under dynamic loading were
obtained and compared with the photogrammetric measurements. The FE model of the
Bent 3 east column was analyzed under the earthquake loading, which were selected to
have biaxial components of 0.5 g maximum ground acceleration in transverse direction
and 0.6 g maximum ground acceleration in the longitudinal direction (Test 4D). The
displacements at grid Point 59, as shown in Fig. 3, were reduced from the
photogrammetric data during Test 4D. These displacements are compared with the
combined components of the displacement transducers mounted on the bridge deck and
the north abutment, which recorded the transverse displacement of Bent 3 at the middle
168
of the bridge deck (DT7), and measured the gap size history and the displacement of the
north abutment in the longitudinal direction of the bridge (DS5 and DS1) during the
same motions. Figure 13 shows the comparison of the displacement history. The FDE
index analysis was completed to compare the results obtained from the photogrammetry
method with the measured displacement transducers. The amplitude error was found to
be 0.10, which represented a very good correlation between the photogrammetry and
displacement transducers results (Table 1). The phase angle error was calculated to be
0.45, which was a good correlation result. Although they were separated by 23 in. (584
mm), the displacements at these two points were similar because of the double curvature
along the column.
The displacement history at the top and bottom hinging regions from the FE
model were also compared with the photogrammetric data. In the bottom hinging region,
the displacement history at Point 2 as shown marked on the grid surface in Fig.3 is
plotted in Fig. 14. Point 51 (Fig. 3) was selected to compare the displacement at the top
hinging region. The periodicities of the two measurements of the motion are very similar
as seen in Figures 14 and 15. The amplitude and phase errors of the lateral displacements
(Table 1) had very good correlations, ranging from 7.3% to 9.4% in amplitude error and
from 14 % to 25% in phase error. The FE model provides very good results.
The effect of yield strength of the longitudinal reinforcement was considered
because limited information about the stress-strain relationships was available to use in
the analysis. The lateral response at the top and bottom points shown in Fig. 3 (Point 2
and 51) was insensitive to the yield strength as shown in Fig. 14 and 15. The vertical
displacement response and cross-sectional rotation were sensitive to yield strength, and
so responses were calculated using yield strength values of 64, 70, 75, and 80 ksi (441,
169
483, 517, and 552 MPa). Figures 16-19 show the comparisons of vertical displacements
at the location of LVDT 3ETR4, which was instrumented 5 in. (127 mm) below the cap
beam as shown in Fig. 3, for each yield strength. Figures 20-27 show the comparisons of
the measured cross-sectional rotations at the locations of LVDTs (LVDT 3ETR3 and
3ETR4 and LVDT 3ETR5 and 3ETR6) that were located at 5 in. (127 mm) and 12 in.
(305 mm) below the cap beam with the calculated FE analysis results. The correlations
between the vertical displacement obtained from the FE model and the LVDT
measurements were examined using the FDE index for each yield strength value as given
in Table 2. As yield strength was increased from 64 ksi to 80 ksi, the total FDE index
decreased from 31% to 18% for vertical displacement calculations. It is evident that the
yield strength of 80 ksi provides a very good comparison as shown in Fig. 28. Figures 29
and 30 show that the FDE index results for the cross-sectional rotations at 5 in. (127 mm)
and 12 in. (305 mm) are very similar to each other for all yield strength values analysis
and represent very good correlations.
The effect of the flexural stiffness of the cap beam was examined to simulate the
rotation of the cap beam during response to ground motions. The flexural stiffness of the
cap beam was calculated and modeled in the program by defining a linear-elastic
rotational spring at the center of the right face of the cap beam. The stiffness was
computed as 10^9 lb-in. (113*10^6 N-m). It was varied from 10^8 lb-in. (113*10^5 Nm) to the real value in the model to determine the effect of the flexural stiffness of the
cap beam on the response of the bridge column. Yield strengths of the reinforcement of
68 and 75 ksi (469 MPa and 517 MPa) were used in the model to find the best definition
of the cap beam stiffness. Figures 31-33 show the comparisons of the vertical
displacements from the analyses with the yield strength defined to be 68 ksi (469 MPa).
170
The stiffness values were taken to be 10^8 lb-in. (113*10^5 N-m), 1.5*10^8 lb-in.
(17*10^6 N-m), and 5*10^8 lb-in. (56.5*10^6 N-m.). The vertical displacement
comparisons are shown in Figs. 34 and 35 for the yield strength of 75 ksi (517 MPa). An
FDE index analysis was used to determine the best correlation between the analyses.
Figures 36 and 37 show that as the stiffness of the cap beam increases, the FDE
decreases. The stiffness of 5*10^8 lb-in. (56.5*10^6 N-m) gives the best comparison
when the yield strength of the steel is equal to 75 ksi (517 MPa) as shown in Table 3. In
Table 3, as the stiffness of the cap beam was increased from 1*10^8 lb-in. (113*10^5 Nm) to 5*10^8 lb-in. (56.5*10^6 N-m), the amplitude error decreased from 41% to
approximately 10%, for either fy=68 ksi (469 MPa) or fy=75 ksi (517 MPa). The phase
error with constant cap beam stiffness increased, however, from 20% to 33% with fy=68
ksi (469 MPa), and from 20% to only 29% with fy=75 ksi (517 MPa).
When reinforced concrete structures are subjected to cyclic loading, they
experience progressive damage and a reduction in stiffness. At very high deformations, it
is inevitable to expect significant damage and stiffness degradation in the bridge column,
therefore, stiffness degradation and stiffness recovery variables were used in the concrete
plastic damage model. The stiffness degradation parameters dc and dt defined in the
concrete plasticity model in ABAQUS are numerical values which depend on inelastic
compressive strain or tensile crack displacement. When the post-peak compressive stress
or tensile crack displacement increases, the material sustains non-recoverable damage,
and the stiffness of the material reduces after a load reversal occurs. In this study, the
compression stiffness degradation parameters for core and cover concrete were defined
as exponential functions dependent on the plastic component of the compressive strain.
Equations 5 and 6 were utilized for core and cover concrete respectively:
171
dc = 1 e
dc = 1 e
30 pl
(5)
160 pl
(6)
where pl is the plastic component of the compressive strain. The compression stiffness
degradation model for the core and cover concrete are shown in Figs. 38 and 39. The
compression stiffness recovery variable (wc) was taken between 0.5 and 1.0, which
implies that as cracks close during load reversal the compression stiffness will be
completely recovered.
The tension stiffness degradation parameter was defined by a linear function in
which a 50 percent reduction of the elastic stiffness was assumed to occur when the
critical crack bandwidth value was reached, and 98 percent of the elastic stiffness when
the critical crack bandwidth was doubled. Figure 40 shows the tension stiffness
degradation model. The tension stiffness recovery variable (wt) was taken to be between
0 and 1.0 in the FE analyses.
Several combinations of different compression and tension stiffness recovery
variables as given in Table 4 were defined in the model to obtain the best parameters.
The comparisons of vertical displacements are shown in Figs. 41-44. The FDE indexes
were also calculated to determine the best combinations of the parameters (Table 4). As
seen in Fig. 45, the best combination of recovery variables are wc =0.8 and wt =1.0.
These stiffness recovery variables were chosen for future analysis.
The effect of the bond-slip at the column interfaces was considered in the FE
analysis because when a column is subjected to axial load and flexure, the longitudinal
reinforcement displaces in tension and in compression as a result of the strain in the
embedded bar. The effects of the slip were modeled using linear-elastic springs at the
beam-column interface. Given the large development length afforded by the foundation
172
and cap beam this was considered to be an adequate assumption. The properties of the
springs were computed as follows. The slip displacement is calculated by integrating the
axial strains over the anchored bar length. The slip of the longitudinal reinforcement
results in additional member end rotation that is not related to flexural deformations.
Experimental results from four double curvature columns indicated that the bar slip
deformations may be as large as the flexural deformations [30]. The value of the slip
deformations did not vary much with increasing lateral displacement or damage in the
tests. Numerous experimental studies have been completed to define the anchorage
behavior of the reinforcing bars. Several analytical bond-slip models have been
developed using these experimental results. Using a bilinear strain distribution as shown
in Fig. 46, the slip in the model was ultimately determined from Eq. 7 and 8.
ld
slip = dx =
s ld
ld '
ld
s y
s ld
l 'd
( s + y )
slip = dx + dx =
+
2
2
0
ld
s > y
(7)
(8)
The development lengths over the elastic and inelastic regions of the
reinforcement (ld and ld) were calculated from the equilibrium of forces in the
reinforcement as given in Eq. 9. A uniform bond stress of ub, which was equal to 6fc,
was assumed in the elastic range (Fig. 46).
ld =
f y db
(9)
4u b
In the inelastic portion of the slip, the slope of the force-slip relationship (Fig. 47)
was assumed to be 1% of the slope of the elastic range. The total slip was calculated to
be 0.08 in. (2 mm) in the FE model having the yield strength of the longitudinal
reinforcement of 64 ksi (441 MPa).
173
The vertical displacement at the location of LVDT 3ETR4 was obtained from the
FE analysis and compared with the measured displacement at the face of the column, as
shown in Fig. 48. The vertical displacement result give higher values than those found in
Fig. 3 when the yield strength was equal to 64 ksi (441 MPa) and no bond-slip behavior
was defined in the FE model. The rotations, however, followed the measured LVDT data
very well as shown in Fig. 49 and 50. An FDE index analysis was used to determine the
effect of the added bond-slip deformation in the FE model (Fig. 51). Table 5 shows that
as the bond-slip effect was introduced in the FE model, the total FDE error increased
from 30% to 43%. At best it can be concluded that bond-slip deformations are not an
important component to the response of the bridge column.
The deformed shapes of the hinging regions were defined over a continuum by
considering the movements of selected Points in the regions [13]. Three grid lines in the
column hinging regions were selected to define the lateral displacements. These lines
were called Middle, Left, and Right lines because of their locations on the grid (Fig. 2).
Four Points on each grid line were selected to obtain the lateral displacements using the
Photogrammetric data. The lateral displacements were connected using a line to simulate
the deformed shape of the column. The lateral displacements of the same points on the
column surface shown in Fig. 3 were also determined from the FE model at the time of
the maximum drift during response to the Test 4D motion. Yield strengths of 64 and 80
ksi (441 and 552 MPa) were considered to obtain results with the FE model. Figure 52
shows the comparisons of the deformed shapes reduced from the photogrammetry data
and the FE model when the yield strength of the steel is equal to 64 ksi (441 MPa). For
the 80 ksi (552 MPa) yield strength, the deformed shapes are compared in Fig. 53. the
deformed shape obtained from the FE model with the yield strength of 80 ksi (552 MPa)
174
follows the deformed shape reduced from the photogrammetry better than the results
with the yield strength of 64 ksi (441 MPa).
4. Conclusions
The best FE model of the reinforced concrete bridge column subjected to

earthquake motion was produced using the following properties:
The static increasing lateral load analysis results showed that using a 1 in.
(25 mm) mesh along the top and bottom hinging regions of the bridge
column provided a similar force-displacement curve as that obtained
using the Modified Compression Field Theory.
According to the analyses results using different yield strengths for the
steel, using the yield strength of 80 ksi (552 MPa) gave the best FDE
index value.
The flexural stiffness of the cap beam should be defined to be 5*10^8 lbin. (56.5*10^6 N-m).
The compression and tension stiffness recovery variables used in the FE

model were determined to be wc =0.8 and wt =1.0.
The error calculated from the vertical displacement response including the
bond-slip effect of the longitudinal reinforcement was higher than the
error associated with the response having no bond-slip effect.
Using these modeling parameters, the FE model adequately represented the drift
response of the bridge column and plastic deformation behavior in response to strong
ground motion. The comparisons of the calculated and measured lateral displacement
175
responses of points in the hinging regions gave very good results, as calculated with the
FDE index analysis. Finally, the deformed shape of the bridge column generated using
the FE model results followed the deformed shape of the column reduced from the
Photogrammetric data.
Table 1 FDE results for lateral displacement comparisons

(FE vs. Photogrammetry measurements)
Phase
Total
error
error
(%)
(%)
10
45
55
7.3
14
22
9.4
25
34
Amplitude
error (%)
Comparison
of Point 59
Comparison
of Point 2
Comparison
of Point 51
176
Table 2 FDE results for different yield strength analyses

(FE vs. LVDT measurements)
Vertical
Displacement
Rotation @ 5
in. below the
cap beam
Rotation @ 12
in. below the
cap beam
Yield
strength of
the steel
(ksi)
64
70
75
80
64
70
75
80
64
70
75
80
Amplitude
error (%)
Phase
error
(%)
Total
error
(%)
8.9
7.3
6.6
6.4
20
20
19
17
7.3
7.3
6.8
6.9
22
17
14
12
6.6
6.6
7.1
9.7
5.4
5.4
5.8
5.9
31
24
21
18
27
27
26
27
13
13
13
13
Table 3 FDE results for analyses with different stiffness of the cap beam analysis
(FE vs. LVDT vertical displacement measurements)
Flexural stiffness
Amplitude
of cap beam lberror (%)
in. (N-m)
fy = 68 ksi (469
MPa)
fy = 75 ksi (517
MPa)
1*10^8 lb-in.
(113*10^5 N-m)
1.5*10^8 lb-in.
(17*10^6 N-m)
5*10^8 lb-in.
(56.5*10^6 N-m)
1*10^8 lb-in.
(113*10^5 N-m)
5*10^8 lb-in.
(56.5*10^6 N-m)
177
Phase
error
(%)
Total
error
(%)
41
20
61
24
23
47
11
33
44
41
20
61
9.4
29
38
Table 4 FDE results for different degradation parameters

(FE vs. vertical deformations 5 in. below tha cap beam as measured with LVDTs)
Phase
error
(%)
Total
error
(%)
11
26
37
11
28
39
11
29
40
8.9
22
31
Stiffness
Amplitude
degredation
error (%)
variables
wc=0.5,
wt=0
wc=0.8,
wt=0
wc=1,
wt=0.8
wc=0.8,
wt=1
Table 5 FDE results for bond-slip model

(FE vs. LVDT vertical displacement measurements)
BondSlip
Model
Amplitude
error (%)
with Slip
without
Slip
Phase
error
(%)
Total
error
(%)
13
31
43
8.9
22
30
282 in. [7.16 m]
348 in. [8.84 m]
348 in. [8.84 m]
282 in. [7.16 m]

14 in. [0.36 m]
SHAKE TABLE
Abutment 1
Bent 1
SHAKE TABLE
Bent 2
SHAKE TABLE
Bent 3
178
Abutment 2
(a)
(b)
Fig. 2 Close view of (a) bottom and (b) top grid systems
DT7
3ETR3
57
56 58
50
49 51
59
52
62
6061
55
5354
43
42 44
45
48
4647
3ETR5
3ETR6
36
35 37
29
28 30
212223
41
3940
34
31 3233
24 252627
38
1617 18
19 20
1112
13
1415
6 7
9 10
1 2
4 5
3EBR5
3EBR7
3ETR4
3EBR6
3EBR8
Fig. 3 Grid system and LVDT locations on column in the Bent 3 east column
179
Lateral Component
0.3
Acceleration (g)
0.2
0.1
0
0
10
20
30
40
-0.1
-0.2
-0.3
Time (sec)
Fig. 4 Lateral component of Northridge earthquake record
Longitudinal Component
0.25
.
0.2
Acceleration (g)
0.15
0.1
0.05
0
-0.05 0
10
20
30
-0.1
-0.15
-0.2
Time (sec)
Fig. 5 Longitudinal component of Northridge earthquake record
180
40
Fig. 6 The aluminum tower and four cameras
m]
in
75.24
14
6 .8
. [3
i
.9 1
50 ]
m
.7 3
75.
.2
. [1
9m
6 7 .8
RC
6 in .
[1 .7 2
m]
24
Tower
2 88
LC
103.8
6
in. [2
.64 m
]
3 6 in
. [ 0 .9
1
in. [7
.32m
28
8
.[
7.
32
.5 3
in
29
m
]
Bent 3
Shake Table
Fig. 7 Location of the Bent 3 east column and the aluminum tower
181
Compression Model (Mander et al. 1984)

7
Confined
6
Unconfined
Stress (ksi)
5
4
3
2
1
0
0
0.02
0.04
0.06
0.08
0.1
0.12
Strain (in./in.)
Fig. 8 Concrete compression model
Tension Model
Bhide_87
Vecchio_82
Collins_87
350
300
Stress (psi)
250
200
150
100
50
0
0
0.005
0.01
Strain
Fig. 9 Concrete tension model
182
0.015
0.02
Steel Model
100000
90000
80000
Stress (psi)
70000
Static
60000
Dynamic
50000
40000
30000
20000
10000
0
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Strain (in./in.)
Fig. 10 Stress-strain relationship for longitudinal reinforcement under monotonic tension

loading
Comparison of Cyclic Behavior

100000
80000
60000
Stress (psi)
40000
20000
0
-0.01
-20000
0.01
0.02
0.03
0.04
0.05
-40000
-60000
Ma et al. (1976)
-80000
FE
-100000
Strain (in./in.)
Fig. 11 Comparison between computed and measured stress-strain response of

reinforcing steel subjected to reversed cyclic loading (Experimental data from Ma et al.
(1976))
183
Monotonic Loading Curve

25
20
MCFT
Force (kips)
15
3 in.
10
Curve
1 in
5
0.5 in
0
0
-5
Displacement (in.)
Fig. 12 Load-deflection curves for monotonic loading analysis of FE model of the bridge
column with various mesh densities
2.5
Photogrammetry
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
Time (sec)
Fig. 13 Comparison between inferred lateral displacements from Photogrammetry

analysis at Point 59 and measured data with displacement transducers
184
Lateral Displacement
Photogrammetry #2
1.00
FE
0.50
Displacement (in.)
0.00
0
10
15
20
25
30
35
40
-0.50
-1.00
-1.50
-2.00
Time (sec)
Fig. 14 Lateral displacement comparison between Point 2 of photogrammetry grid and

FE analysis at the bottom hinging region
Lateral Displacement
Photogrammetry #51
1.50
Displacement (in.)
FE
1.00
0.50
0.00
5
10
15
20
25
30
35
40
-0.50
-1.00
-1.50
-2.00
Time (sec)
Fig. 15 Lateral displacement comparison between Point 51 of photogrammetry grid and

FE analysis at the top hinging region
185
Vertical Disp. @Column Face

LVDT 3ETR4
0.12
FE @5in top
0.1
0.08
D isp. (in.)
0.06
0.04
0.02
0
-0.02 5
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
Fig. 16 Vertical displacement comparison between FE analysis result and LVDT

measurements at the location of LVDT 3ETR4 (fy=64 ksi)*
Vertical Disp.@Column Face

LVDT 3ETR4
FE @5in top
0.1
0.08
0.06
D isp. (in.)
0.04
0.02
0
-0.02
10
15
20
25
30
35
-0.04
-0.06
-0.08
Time (sec)

measurements at the location of LVDT 3ETR4 (fy=70 ksi) *
Infinitely stiff cap beam, wc =0.8 and wt =1.0.
186
40
Vertical Disp.@Column Face

LVDT 3ETR4
FE @5in top
0.08
0.06
Disp. (in.)
0.04
0.02
0
-0.02
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)

LVDT 3ETR4
0.08
FE @5 in top
0.06
Disp. (in.)
0.04
0.02
0
-0.02 5
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)

*

187
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 20 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR3 and 3ETR4 (fy=64 ksi)*
0.01
FE@10
LVDT@12
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
measurements at the location of LVDT 3ETR5 and 3ETR6 (fy=64 ksi) *
*

188
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
FE@10
0.01
LVDT@12
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
*
189
FE @5 in.
0.01
LVDT @ 5 in.
0.005
0
Rotation (rad)
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
FE@10
0.01
LVDT@12
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
*

190
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
Time (sec)
FE @10 in.
0.01
LVDT @ 12
in.
0.005
0
Rotation (rad)
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
*

191
Amplitude Error
35
Phase Error
30
Total Error
Error (%)
25
20
15
10
5
0
64
70
75
80
Yield strength of the reinforcement (ksi)
Fig. 28 The comparison of FDE index errors for different yield strength of steel (FE vs.
LVDT vertical displacement measurements)
Cross-sectional Rotation @ 5 in.
Amplitude Error
Phase Error
Total Error
35
30
Error (%)
25
20
15
10
5
0
64
70
75
80
Fig. 29 The comparison of FDE index errors for cross-sectional rotation at 5 in. below
cap beam using different yield strength of steel (FE vs. LVDT vertical displacement
measurements)
192
Cross-sectional Rotation @ 12 in.

Amplitude Error
35
Phase Error
Total Error
30
Error (%)
25
20
15
10
5
0
64
70
75
80
Fig. 30 The comparison of FDE index errors for cross-sectional rotation at 12 in. below
cap beam using different yield strength of steel (FE vs. LVDT vertical displacement
measurements)
LVDT 3ETR4
Disp. (in.)
0.06
FE @5in top
0.04
0.02
0
-0.02
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)

measurements at the location of LVDT 3ETR4 for K=1*10^8 lb-in. (fy=68 ksi)*
*
Linear-elastic rotational spring, wc =0.8 and wt =1.0.

193

LVDT 3ETR4
Disp. (in.)
0.06
FE @5in top
0.04
0.02
0
0
10
15
20
25
30
35
40
-0.02
-0.04
-0.06
-0.08
Time (sec)

measurements for K=1.5*10^8 lb-in. (fy=68 ksi) *
LVDT 3ETR4
0.15
.
FE @5in top
Disp. (in.)
0.1
0.05
0
0
10
15
20
25
30
35
40
-0.05
-0.1
Time (sec)

measurements for K=5*10^8 lb-in. (fy=68 ksi) *
*
Linear-elastic rotational spring with 10^8 lb-in. capacity, wc =0.8 and wt =1.0.
194

LVDT 3ETR4
0.06
.
FE @5in top
Disp. (in.)
0.04
0.02
0
-0.02
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)

LVDT 3ETR4
FE @5in top
0.12
0.1
Disp. (in.)
0.08
0.06
0.04
0.02
0
-0.02 0
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)

*
195
Yield Strength= 68 ksi (469 MPa)

Amplitude error
Phase error
70
Total error
60
Error (%)
50
40
30
20
10
0
1*10^8 lb-in. (113*10^5
N-m)
1.5*10^8 lb-in. (17*10^6 5*10^8 lb-in. (56.5*10^6

N-m)
N-m)
Stiffness of the cap beam
Fig. 36 The comparison of FDE index errors for different stiffness of the cap beam (FE
vs. LVDT vertical displacement measurements) (fy=68 ksi) *
196
Yield Strength= 75 ksi (517 MPa)

Amplitude error
Phase error
70
Total error
60
Error (%)
50
40
30
20
10
0
1*10^8 lb-in. (113*10^5 N-m)
5*10^8 lb-in. (56.5*10^6 N-m)
Stiffness of the cap beam
Fig. 37 The comparison of FDE index errors for different stiffness of the cap beam (FE
vs. LVDT vertical displacement measurements) (fy=75 ksi) *

197
Core Concrete Damage Degredation Model

1.2
1
1-dc
0.8
0.6
0.4
0.2
0
0
0.02
0.04
0.06
0.08
0.1
0.12
inelastic strain (in./in.)
Fig. 38 The compression stiffness degradation model for core concrete of the bridge
column
Cover Concrete Damage Degredation Model
1.2
1
1-dc
0.8
0.6
0.4
0.2
0
0
0.005
0.01
0.015
0.02
inelastic strain (in./in.)
Fig. 39 The compression stiffness degradation model for cover concrete of the bridge
column
198
Bhide Tension Stiffness Degredation Model

1.2
1
1-dt
0.8
0.6
0.4
0.2
0
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
Cracking Disp. (in.)
Fig. 40 The tension stiffness degradation model for core and cover concrete of the bridge
column
LVDT 3ETR4
FE @5in top
0.15
D isp. (in.)
0.1
0.05
0
5
10
15
20
25
30
35
-0.05
-0.1
Time (sec)

measurements for wc=0.5, wt= 0*
*
Infinitely stiff cap beam.
199
40

LVDT 3ETR4
0.15
FE @5in top
D isp . (in .)
0.1
0.05
0
5
10
15
20
25
30
35
40
-0.05
-0.1
Time (sec)

measurements for wc=0.8, wt= 0 *

LVDT 3ETR4
FE @5in top
0.15
Disp. (in.)
0.1
0.05
0
5
10
15
20
25
30
35
-0.05
-0.1
Time (sec)

measurements for wc=1, wt= 0.8*
*
200
40

LVDT 3ETR4
0.12
FE @5in top
0.1
0.08
Disp. (in.)
0.06
0.04
0.02
0
-0.02 5
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)

measurements for wc=0.8, wt= 1*
50
Amplitude error
Phase error
45
Total error
40
Error (%)
35
30
25
20
15
10
5
0
wc=0.5 wt=0
wc=0.8 wt=0
wc=1 wt=0.8
wc=0.8 wt=1
Recovery and degredation parameters
Fig. 45 The comparison of FDE index errors for different recovery and degradation
parameters (FE vs. LVDT vertical displacement measurements) *
*

201
Fig. 46 Column deformations due to bar slip
12000
Force (lb)
10000
8000
6000
4000
2000
0
0
0.02
0.04
0.06
0.08
0.1
Slip (in.)
Fig. 47 Assumed slip behavior in the FE model
202

LVDT 3ETR4
FE @5in top
0.2
Disp. (in.)
0.15
0.1
0.05
0
5
10
15
20
25
30
35
40
-0.05
-0.1
Time (sec)

measurements at the location of LVDT 3ETR4*
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 49 The cross-sectional rotation comparison at the location of LVDT 3ETR3 and
3ETR4*
*
Infinitely stiff cap beam, bond-slip effect included at the interfaces, wc=0.8, wt= 1.
203
0.015
FE@10
LVDT@12
0.01
Rotation (rad)
0.005
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 50 The cross-sectional rotation comparison at the location of LVDT 3ETR5 and
3ETR6*
Yield Strength= 64 ksi (441 Mpa)
Amplitude error
Phase error
70
Total error
60
Error (%)
50
40
30
20
10
0
with Slip
without Slip
Fig. 51 The comparison of FDE index errors for bond-slip models (FE vs. LVDT vertical
displacement measurements)*
*
204
Bent 3 East Column Deformed Shape at Peak Displacement

(Test4D)
80
Middle Line_Photogrammetry
Middle Line_FEA
70
Left Line_Photogrammetry
Left Line_FEA
60
Right Line_Photogrammetry
Right Line_FEA
Column Height (in.)
50
40
30
20
10
0
-0.2
-0.1
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Lateral Disp. (in.)
Fig. 52 Deformed shape of the bridge column at the maximum drift during Test 4D
(fy=64 ksi)*
205
Bent 3 East Column Deformed Shape at Peak Displacement

(Test4D)
80
Middle
Line_Photogrammetry
Middle Line_FEA
Left
Line_Photogrammetry
Left Line_FEA
60
Right
Line_Photogrammetry
Right Line_FEA
70
Column Height (in.)
50
40
30
20
10
0
-0.2
-0.1
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Lateral Disp. (in.)
Fig. 53 Deformed shape of the bridge column at the maximum drift during Test 4D
(fy=80 ksi) *
206
REFERENCES
1. Baker A.L.L. Ultimate load theory applied to the design of reinforced and
prestressed concrete frames. Concrete Publications Ltd., London. 1956.
2. Baker A.L.L., Amarakone A.M.N. Inelastic hyperstatic frame analysis. Proc.
International Symposium on the Flexural Mechanics of Reinforced Concrete,
ACI SP-12 Miami, pp. 85-142. 1964.
3. Mattock A.H. Rotational capacity of hinging regions in reinforced concrete
beams. Proc. International Symposium on the Flexural Mechanics of Reinforced
Concrete, ACI SP-12 Miami, pp. 143-181. 1964.
4. Mattock, A.H. Discussion of rotational capacity of hinging regions in reinforced
concrete beams. by Corley, W.G., Journal of the Structural Division, ASCE, V.
93, No. ST2, Apr., pp. 519-522. 1967.
5. Corley W. G. Rotational capacity of reinforced concrete beams. J Struct Division
ASCE 1966; V. 92, No. ST5, pp. 121-146.
6. Park R., Priestley M. J. N., Gill W. D. Ductility of square-confined concrete
columns. J Struct ASCE 1982; 108 No. ST4. 929-950.
7. Priestley M. J. N., Park R. Strength and ductility of concrete bridge columns
under seismic loading. ACI Struct J 1987 V.84 Issue 1. pp. 61-76.
8. Paulay T., Priestley M. J. N. Seismic design of reinforced concrete and masonry
structures. John Wiley & Sons, Inc. 1992.
9. Sheikh S.A., Khoury S.S. Confined concrete columns with stubs. ACI Struct J
1993 V. 90 No. 4 pp. 414-431.
207
10. Mendis P. Plastic hinge lengths of normal and high-strength concrete in flexure.
Advances in Structural Engineering 2001 V. 4 No. 4 pp. 189-195.
11. Firat
Alemdar
Z.,
Browning
J.,
Olafsen
J.
2010.
Photogrammetric
Measurements of RC Bridge Column Deformations, Engineering Structures,

submitted.
12. ABAQUS, Version 6.8-2, Simulia, 2009, http://www.simulia.com.
13. Firat Alemdar Z., Browning J., Olafsen J., and Nick Hunt. 2008. Monitoring RC
Bridge Column Hinging With Photogrammetry, 14th World Conference on
Earthquake Engineering, Bejing, China.
14. Mander J. B., Priestley M. J. N., and Park R. Seismic Design of Bridge Piers.
Department of Civil Engineering Research Report 84-2, University of
Canterbury, 483 pp. 1984.
15. Scott, B.D., Park, R., and Priestley, M.J.N. Stress-strain behavior of concrete
confined by overlapping hoops at low and high strain rates. ACI Struct J 1982 V.
79 No. 1 pp. 13-27.
16. Spacone, E., Filippou, F., and Taucer, F. F. Fibre beam-column model for
nonlinear analysis of r/c frames: part 1. formulation. Earthquake Eng. And
Structural Dynamics, V. 25, 711-725, 1996.
17. Yassin, M. Nonlinear analysis of prestressed concrete structures under monotonic
and cyclic loads. Ph.D. Thesis, University of California, Berkeley 1994.
18. Watstein, D. Effect of straining rate on the compressive strength and elastic
properties of concrete. Am. Conc. Inst. J., 24(8), 729-744. 1953.
19. Wakabayashi,
M.
Design
of
Earthquake-Resistant
York:McGraw-Hill, 1986.
208
Buildings.
New
20. Hosoya, H. et al. Strain rate in the members of reinforced concrete frame
structure during an earthquake and its effects. Journal of Structural and
Construction Engineering (Transactions of AIJ) 499 (1997): 77-83.
21. Mahin, S.A., V.V. Bertero, M.B. Atalya and D. Rea. Rate of loading effects on
uncracked and repaired reinforced concrete members. Report UCB/EERC-72/09.
Berkeley: EERC, University of California, 1972.
22. Bhide, S. B., and Collins, M. P. Reinforced concrete elements in shear and
tension. Publication No. 87-02, Department of Civil Engineering, University of
Toronto, Jan. 1987, 308 pp.
23. Vecchio, F. J., and Collins, M. P. Response of reinforced concrete to in-plane
shear and normal stresses. Publ. No. 82-03, Dept. of Civil Engineering, Univ. of
Toronto. 1982.
24. Collins, M. P., and Mitchell, D.
Prestressed concrete basics. Canadian
Prestressed Concrete Institute, Ottawa. 1987.

25. ACI Committee 439. Effect of steel strength and of reinforcement ratio on the
mode of failure and strain energy capacity of reinforced concrete beams. ACI
Journal 66 (3) (1969): 165-173.
26. CEB-FIP. CEB-FIP Model Code 1990: Design Code. London: T Telford, 1993.
27. Ma, S-Y.M., V.V. Bertero and E.P. Popov. Experimental and analytical studies of
the hysteretic behavior of reinforced concrete rectangular and t-beams. Report
No. EERC-76-2. Berkeley: EERC, University of California, 1976.
28. Response-2000. Reinforced concrete sectional analysis using the modified
compression
field
theory.
Version
http://www.ecf.utoronto.ca/~bentz/r2k.htm.
209
1.0.5,
last
accessed
2010,
29. Dragovich JJ., Lepage A. FDE index for goodness-of-fit between measured and
calculated response signals. Earthquake Engineering and Structural Dynamics
2009;38:1751-1758.
30. Sezen, H. Seismic Behavior and Modeling of Reinforced Concrete Building
Columns. Ph.D. Thesis. University of California, Berkeley. 2002.
210
MANUSCRIPT 3: PLASTIC HINGE LENGTH EXPRESSION

FOR RC BRIDGE COLUMNS
a,*
Zeynep Firat Alemdar , JoAnn Browningb and Adolfo Matamorosc
USA
USA
USA
Corresponding author Tel: +1 785 864 3760 Fax: +1 785 864 5631 E-mail addresses:
zeynep@ku.edu
(Z. F. Alemdar).
1. Introduction
The determination of magnitude and location of inelastic deformations in
reinforced concrete bridge columns is a critical step for characterizing the performance
of the bridge system in earthquake events. Although it is possible to believe that some
ductility will be provided by beam hinges in bridge systems, it is generally the columns
of the bridges that must have inelastic rotational capacity. Bridge systems are designed to
keep inelastic behavior in the columns and away from the superstructure, which is
different than building systems. Plastic hinging regions indicate the area of concentrated
damage for bridge columns that experience inelastic deformations. Therefore, a
consistent prediction of a plastic hinge length is also necessary to evaluate the length of
the column that needs to be confined along the critical section. Simple expressions for
211
plastic hinging regions also enable the development of estimates of column drift capacity
and drift demand.
The plastic hinge length in reinforced concrete columns has been studied widely
by many researchers [2-16]. Previous experimental research has generally been
conducted using static loading and for small-scale components such as beams and
columns. The real load distribution due to earthquake excitations of a structure would be
different from static loading, and elements would behave differently. In addition, an
evaluation of large-scale system behavior will include the effects of moment
redistribution and progression of yielding throughout the entire structure.
Only a few of the previous studies have focused on determining the plastic hinge
length in reinforced concrete bridge columns, and instead have focused on building-type
columns. The location and the progression of plastic hinge length in reinforced concrete
bridge columns would differ from reinforced concrete buildings. In addition, many of the
parameters such as axial load ratio, diameter of the longitudinal reinforcement and the
cross-sectional dimensions of the columns in the previous tests are not appropriate for
reinforced concrete bridge columns [12-16, 24].
Expressions that have been developed to estimate the length of plastic hinges
have either been based on the maximum drift at the top of the column, or the spread of
plastic behavior in the hinging regions. Plastic hinge lengths have been calculated for 51
of the reported tests from the measured maximum top displacement using the lumpedplasticity model [7, 10, 17, 20, 21, 26-29]. Only 21 test data included the measured
plastic hinge length from using the spread of plasticity approach [15, 16, 19, 22, 23, and
25]. The lumped-plasticity model has generally been used to simulate the nonlinear
response of bridge columns and estimate the ultimate displacement capacity, which is an
212
important requirement of performance-based earthquake engineering (PBEE). In

addition, Mander (1983) stated that the plastic hinge length must not be used for
estimating the length requiring detailed confinement because plastic curvature would
spread over approximately three equivalent plastic hinge lengths. Therefore, the
evaluation of the spread of plasticity in the hinging regions might be useful to predict the
required detailing in the confinement region. It has also observed that the equivalent
plastic hinge length and the region of plasticity where special reinforcing detailing is
required must be defined separately to ensure dependable inelastic rotation capacity [11].
This research defines a new plastic hinge length expression that is appropriate for
static and dynamic loading cases for reinforced concrete bridge columns. Previous
plastic hinge length expressions were evaluated using the static test measurements. In the
new expression, however, both static and dynamic test results have been considered to
obtain the best fit for every loading condition. The new expression has been obtained by
doing a multivariable linear regression analysis using the previous test results in the
literature.
The drift capacity of a bridge column, which corresponds to a 20% reduction in
lateral load capacity on the descending branch of the response backbone curve, is
estimated using the new expression and compared with the results that were obtained
from the earlier plastic hinge length expressions. An expression to calculate the
maximum drift capacity of a bridge column in double curvature is derived by considering
the deformations due to flexure as influenced by the definition of plastic hinge length
(lp), and the bond-slip effect of the longitudinal reinforcement at the connections. A fourspan large-scale bridge system was tested under biaxial earthquake loadings at the
University of Nevada Reno, as shown in Fig. 1, to obtain the transverse and longitudinal
213
displacements at the cap beam on the bridge columns of Bent 3 and compared with the
calculated response from the expression [30].
2.
Analytical Investigation: Expressions Derived Using Static Test Results

Previous plastic hinge length models were developed using experimental results
from static tests. A suite of 115 column tests that represent typical elements in reinforced
concrete buildings and bridges that encompass a wide range of parameters were used.
Axial load level, compressive strength of concrete, yield strength of longitudinal
reinforcement, diameter of longitudinal reinforcement, and section dimensions are some
of the main variables in the static tests. The minimum and maximum ranges of the main
variables used in the literature are given in Table 1.
In the studies performed before 1983, the cross-section dimensions of a column
were the main parameters used to determine the plastic hinge length of a column. With
the results of Manders tests at the University of Canterbury [8], the strain penetration
effect of the longitudinal reinforcement was found to be an important factor within the
plastic hinge region, and the diameter of the longitudinal reinforcement was incorporated
into new expressions. After the effect of axial load on plastic hinge length calculation
was discovered, Soesianawati et al. (1986) and Watson et al. (1994) added the axial load
ratio into their expressions. Later expressions incorporate dimensionless parameters such
as longitudinal reinforcement ratio, and shear span to depth ratio into expressions [15].
Table 2 shows various plastic hinge length expressions that have been developed to
predict the plastic hinge length of reinforced concrete columns.
A small suite of 10 dynamic test results with relevance to determining plastic
hinge length expressions have been collected. Hachem et al. (2003) and Dodd et al.
214
(2000) used the equation proposed by Priestley et al. (1992) and estimated the plastic
hinge length before testing circular bridge columns under uniaxial and bidirectional
earthquake loading. The main parameters in the tests were aspect ratio of the columns (4
to 10) to characterize the flexurally dominated columns, axial load level (0.04Agfc 0.4
Agfc), and longitudinal reinforcement ratio (1.2%-1.62%). The plastic hinge length was
measured according to the spread of plasticity model in the study of Hachem et al.
(2003). Dodd et al. (2000) calculated the plastic hinge lengths from the maximum drift of
the columns. After testing the columns, it was also observed that the estimated plastic
hinge lengths were within the range of measured values as shown in Table 3. The
average measured plastic hinge length value from the testing of four identical columns
[19] was about 13 in. (330 mm) as given in PEER 2003/06 report. It was clear from the
experimental values that the equation proposed by Priestley et al. (1992) overestimated
the plastic hinge length for all of the columns as shown in Table 3.
Two new expressions to calculate the plastic hinge length in reinforced concrete
bridge columns have been derived using the measured plastic hinge length values from
yield curvature and the plastic hinge length values that were calibrated to calculate the
correct maximum top displacement during the static and dynamic tests. The experimental
data was limited to contain the data which characterize modern bridge column design
criteria as described next, and included 72 out of the 115 tests from the literature. Fifty
one plastic hinge length data included the plastic hinge length values calculated from the
measured top displacement as given in Table 4. The remaining 21 test data measured the
plastic hinge length as defined where the curvature is bigger than the yield curvature of
the columns that were tested, as shown in Table 5. The test data was limited to represent
modern bridge columns, which meet the following design criteria:
215
An axial load ratio P/fcAg less than or equal to 0.3, where P is the column
axial load, fc is the concrete compressive strength, Ag is the gross section
area of the column,
A longitudinal reinforcement ratio less than or equal to 3%,
A compressive strength of the concrete less than or equal to 10 ksi to

represent modern bridge columns.
Each parameter of the equations in Table 1.2 was considered in the analysis to
determine their affect on estimates of plastic hinge lengths. In order to obtain the best fit
expressions for the plastic hinge length over the cross-sectional dimension, multivariable
linear regression analysis was used. A number of different combinations of parameters
were examined including the length of the columns, axial load ratio, diameter of the
longitudinal reinforcement, yield strength of the reinforcement, compressive strength of
the concrete, shear span to depth ratio, and the diameter of the longitudinal reinforcement
normalized with yield strength of the reinforcement and the compressive strength of the
concrete. The determining factor for selecting the best-fit equation was the minimum
coefficient of variation (COV) of measured Lp/d (ratio of plastic hinge length to section
dimension) to that of calculated Lp/d. After analyses of the various combinations, the
column height and the diameter of the longitudinal reinforcement normalized with yield
strength of the reinforcement and the compressive strength of the concrete (representing
bending along the column height and the bond-slip deformation effect) were found to be
the parameters that produced the best-fit expressions. The best-fit expression derived
using 51 test data based on calibrated value of lp to calculate the maximum drift capacity
is given in Eq. 1.
216
Lp
d
Lp
d
1 3 f y db
L
+
+
5 500 f c' 1000
(psi and in.)
(1a)
3 f y db
1
L
+
+
'
4 10000 f c 25000
(MPa and mm)
(1b)
The best-fit expression based on measured spread of plasticity (21 test data set) is given
in Eq. 2.
Lp
d
Lp
d
f y db
3
L
+
+
(psi and in.)
10 1000 f c' 2500
(2a)
3 f y db
3
L
+
+
10 100000 f c' 5000
(2b)
(MPa and mm)
The new expressions were compared with previous plastic hinge length models.
In order to compare the different expressions, the measured plastic hinge length was
divided by the calculated plastic hinge length obtained from the expressions available.
The best fit lines were drawn using the ratios of measured to calculated plastic hinge
lengths based on each equation available in the literature. Figure 2 shows the
comparisons of the ratio of measured to calculated plastic hinge lengths with varied
parameter (fydb/fc) in Eq. 1 for all data. The ratio of measured to calculated plastic
hinge length results were closest to unity for the new expression. The ratio of measured
to calculated plastic hinge length values determined by the equations of Park (1982) and
Paulay (1992) were the next best for representing the data at low axial loads. The
comparison of the ratio of measured to calculated plastic hinge length result with varied
bridge column height is shown in Fig. 3 for Eq. 1. The ratio of measured to calculated
plastic hinge length results from the new expression (Eq. 1) were the closest to unity
with respect to the column height. The ratio of measured to calculated plastic hinge
length values determined by the equations of Mander (1983), Park (1987), and Paulay
217
(1992) intersected unity well, which was very similar to the results found in Fig. 1 except
Paulay (1992) overestimated the plastic hinge length with respect to column height.
The calculated COV values for the ratio of measured to calculated plastic hinge
length are between 0.37 and 0.57 for each expression. The new expression had the lowest
COV value equal to 0.37. The COV of Park et al. (1982) equation was 0.39, which is the
second best COV value after the new expression. Baes equation gave the highest COV
(0.57) for all data with low axial load. The COV values were 0.41, 0.42, 0.45, and 0.45
for Paulay (1992), Park (1987), Mander (1983), and Berry (2008) respectively.
Figure 4 shows the comparisons of the ratio of measured to calculated plastic
hinge lengths with varied value of the first parameter in Equation 2 and the best fit lines
drawn using the ratios for each available expression based on the 21 test data set. The
ratio of measured to calculated plastic hinge length results were closest to unity for the
new expression with respect to fydb/fc. The ratio of measured to calculated plastic hinge
length values determined by the equations of Mander (1983) and Park (1987) were the
next best results after the new expression values. The comparison of the ratio of
measured to calculated plastic hinge lengths with varied bridge column height is shown
in Fig. 5 for the 21 test data set. The ratio of measured to calculated plastic hinge length
results from the new expression (Eq. 2) were the closest to unity with respect to the
column height. The ratio of measured to calculated plastic hinge length values
determined by the equations of Mander (1983) and Park (1987) followed a similar trend
as the new expression.
The calculated COV values for each expression using the analysis of 21 test data
based on the measured spread of plasticity are between 0.27 and 0.46. The new
expression had a COV value equal to 0.30, and the COV of the Mander (1983) equation
218
was 0.27. The Bae (2005) and Berry (2008) equations gave the highest COV (0.46) for
low axial load level. The COV values were 0.32, 0.40, and 0.44 for Park (1987), Park
(1982), and Paulay (1992) respectively.
The COV of Eq. 1 was also determined using the 21 test data of plastic hinge
length derived considering the length of the column with curvature values greater than
the yield curvature (>y) to determine how reliable Eq. 1 is to estimate the actual spread
of plasticity. The ratios of plastic hinge length values from the new expression (Eq. 1) to
the measured plastic hinge length results were calculated. The COV was found to be
0.34, which is an average COV compared to that of the other expressions. This analysis
shows that the behavior of columns when considering the maximum drift capacity and
spread of plasticity (>y) is different.
It is of interest to consider the effect of the axial load level when using each
expression. All the expressions except Berry (2008), which was derived only for bridge
columns with P/Agfc less than or equal to 0.3, were developed for a wide range of axial
loads. In the literature, some of the plastic hinge length expressions work well for high
axial load ratios, whereas they fail to give satisfactory results for low axial load ratios.
Figure 6 shows the effectiveness of the plastic hinge length calculations for a wide range
of axial load ratios. Although only Baes equation includes an axial load parameter, some
of the other equations show considerable changes of measured to calculated ratios of
Lp/d with increasing levels of axial load. It is important to note that the COV values
varied between 0.16 and 0.30 for high axial load ratios. The COV values were 0.16, 0.17,
0.17, 0.18, 0.27, 0.28, 0.28, and 0.30 for Park (1987), Mander (1983), Berry (2008),
Priestley (1992), Bae (2005), Park et al. (1982), and the proposed equations (1.1 and 1.2)
respectively. It appears that most equations have a much better correlation for the plastic
219
hinge length with high axial load, and the behavior of columns is quite different under
high axial load than low axial load. The proposed expression had a good COV value
equal to 0.30, which compared very well to the COV at low axial load ratio of (0.37).
Therefore, the use of Eq. 1 and Eq. 2 provide satisfactory predictions both for high and
low axial load levels. At high axial load ratios, the proposed expression can be
normalized by 1.4 to provide a good estimate for the plastic hinge length.
3.
Analytical Investigation: Correlation With Measured Maximum Drift

The new expression was used to calculate the maximum drift of the bridge
column tested at UNR as described previously. A drift expression was derived using a
lumped-plasticity model with a fixed plastic hinge length (lp) for a double curvature
column. Flexural deformations and slip of the longitudinal reinforcement at the cap beam
and footing interface were considered in the maximum drift expression, and given in
Equation 1.3. The first and last terms in Eq. 3 represent the flexural displacement in the
elastic and plastic range, and the second term represents the slip contribution to the top
displacement.
lp
d
yl 2
6
db f y
48 E s
*
f 'c
L
+ pl p (L l p )
d d'
(3)
The bridge column was tested using the biaxial applications of scaled motions
recorded at the Century City Country Club during the 1994 Northridge, California
earthquake [31]. The maximum drift was measured to be 5.3 in. in the transverse
direction of the bridge after all the motions were applied. The maximum drift values of
the bridge column were calculated using the plastic hinge length expressions (Table 2)
and compared with the measured data in Table 5. The Park (1982) plastic hinge length
220
equation gave the closest value (5.94 in.) to the measured test data. The proposed
expression (Eq. 1) also estimated the measured data very well and provided the second
best estimate of 6.5 in. after the Park (1982) equation. The maximum drift value was also
determined using Equation 2 and found to be 7.1 in., which is the third best correlation to
the real test data.
4.
Conclusions
A review of literature shows that an expression for plastic hinge length can be
developed based on values used to produce best estimates of maximum drift capacity or
can be developed to represent the spread of plasticity along the height of the column.
This study produced equations that characterize the behavior of these two groups of data
better than any existing equation of lp, and shows that the behavior is different between
these two methods. The following conclusions are made:
1. Estimates of plastic hinge length used to calculate the maximum drift
capacity are shorter in value than estimates of plastic hinge length
based on spread of plastic curvature over height of column.
2. The spread of plasticity at low axial loads is different than at high
axial loads. Most existing equations are better fit at high axial loads.
The proposed equations have nearly constant fit across axial loads,
but best fit at low axial loads, which represent bridge columns.
3. The maximum drift of the one column tested under dynamic loads
was best represented using the simple plastic hinge length estimate
proposed by Park (1982). More parametric study is needed to
understand the spread of plasticity.
221
Table 1 Static test data properties

L, in.
Parameters
db, in.
fc, ksi
fy, ksi
2
P/Pu
(mm)
(mm)
(%)
(kN/mm )
(kN/mm2)
(1)
(2)
(3)
(4)
(5)
(6)
0.24 (6.1)
0.7
3.1 (0.021)
43.5 (0.3)
16.25
75.55
(0.11)
(0.52)
Minimum
0.003 25 (635)
Maximum
0.77
360
1.75
(9144)
(44.45)
5.65
1: Axial load ratio

2: The length of column
4: Longitudinal reinforcement ratio
5: The compressive strength of concrete
Table 2 Expressions available in the literature

Authors
Expressions
Park et al. (1982), ksi

(MPa)
0.4*d (1)
Park (1987), ksi (MPa)
0.08*L (2)+6db (3) (0.08*L+0.88*db)
Priestley (1992), ksi (MPa)
0.08*L+.15dblfye (4)>0.3dblfye
(0.08*L+.022dblfye>0.3dblfye)
Mander (1983), ksi (MPa)

Bae (2005)
6.35*(db)+0.06*L
(32*(db)+0.06*L)
(5)
{(0.3*(P/P0 )+3*(As (6)/Ag)0.1)*(L/h)+.25}*h
Berry (2008), psi (MPa)
optimal: 0.0375*L+0.01*fy (7)db/sqrt(fc')(8)

(0.0375*L+0.12*fydb/sqrt(f'c) )
simplified: 0.05*L+.008fydb/sqrt(fc')
(0.05*L+0.1fydb/sqrt(f'c))
1: The section depth

2: The length of column
5: Axial load capacity of the section
6: Area of the longitudinal reinforcement
8: Compressive strength of the concrete
222
Table 3 Calculated and measured plastic hinge length results.
Hachem
(2003)
Dodd
(2000)
L/d(1)
Calculated
Lp, in. (mm)
Measured
Lp, in.
(mm)
13.30 (338)
13 (330)
4.84 (123)
3.54 (90)
4.84 (123)
3.35 (85)
6.73 (171)
4.92 (125)
6.73 (171)
5.31 (135)
10
8.62 (219)
7.68 (195)
10
8.62 (219)
5.91 (150)
4.84 (123)
3.35 (85)
6.73 (171)
5.31 (135)
4.84 (123)
3.74 (95)
1: Aspect ratio
223
Table 4 Column properties and calculated plastic hinge length values
Park
(1982)
Priestley
(1987)
Tanaka
(1990)
Kovacic
(1995)
d or
h, in.
(mm)
(1)
22
(559)
22
(559)
19.69
(500)
19.69
(500)
19.69
(500)
23.62
(600)
15.75
(400)
15.75
(400)
29.53
(750)
29.53
(750)
29.53
(750)
15.75
(400)
15.75
(400)
15.75
(400)
15.75
(400)
21.65
(550)
21.65
(550)
21.65
(550)
21.65
(550)
6.46
(164)
6.46
(164)
6.46
(164)
6.46
(164)
Measured
Lp, in.
(mm)
(2)
9.68
(246)
7.48
(190)
10.63
(270)
11.42
(290)
8.86
(225)
8.27
(210)
8.51
(216)
8.66
(220)
10.93
(278)
11.81
(300)
12.31
(313)
6.77
(172)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
0.44
0.26
2.18
0.34
0.21
2.18
0.54
0.11
4.00
0.58
0.11
4.00
0.45
0.00
5.50
0.35
0.15
2.00
0.54
0.20
4.00
0.55
0.21
4.00
0.37
0.10
4.27
0.40
0.30
4.27
0.42
0.30
4.27
0.43
0.20
4.00
0.55
0.20
4.00
0.67
0.20
4.00
0.70
0.20
4.00
0.39
0.10
3.00
0.63
0.10
3.00
0.67
0.30
3.00
0.86
0.30
3.00
1.5 (38)
0.25
0.16
8.46
2.76 (70)
0.47
0.17
8.46
3.74 (95)
0.63
0.14
8.46
1.77 (45)
0.30
0.19
8.46
8.7 (221)
10.6
(269)
11.06
(281)
8.35
(212)
13.66
(347)
14.45
(367)
18.66
(474)
224
L, in.
(mm)
(6)
48
(1219)
48
(1219)
70
(1778)
108
(2743)
108
(2743)
47
(1194)
63
(1600)
63
(1600)
126
(3200)
126
(3200)
126
(3200)
63
(1600)
63
(1600)
63
(1600)
63
(1600)
65
(1651)
65
(1651)
65
(1651)
65
(1651)
50
(1270)
50
(1270)
50
(1270)
50
(1270)
db,
in.
(mm)
(7)
0.94
(24)
0.94
(24)
0.51
(13)
0.51
(13)
0.51
(13)
0.94
(24)
0.63
(16)
0.63
(16)
0.39
(10)
0.39
(10)
0.39
(10)
0.79
(20)
0.79
(20)
0.79
(20)
0.79
(20)
0.79
(20)
0.5
(13)
0.5
(13)
0.5
(13)
0.47
(12)
0.47
(12)
0.47
(12)
0.47
(12)
fy, ksi
(MPa)
(8)
55.1
(380)
55.1
(380)
54.08
(373)
53.8
(371)
44.22
(305)
43.94
(303)
44.67
(308)
61.93
(427)
48.58
(335)
48.58
(335)
48.58
(335)
68.73
(474)
74.1
(511)
74.1
(511)
74.1
(511)
74.1
(511)
68
(469)
68
(469)
68
(469)
58
(400)
58
(400)
58
(400)
58
(400)
f'c,
ksi
(MPa)
(9)
3.35
(23)
6
(41)
4.81
(33)
5.04
(35)
5.8
(40)
4.35
(30)
3.77
(26)
3.63
(25)
5.8
(40)
4.21
(29)
4.64
(32)
3.71
(26)
4.64
(32)
4.64
(32)
4.65
(32)
4.65
(32)
l
(%)
(10)
1.79
1.79
2.70
2.70
2.57
2.43
2.43
1.51
2.84
2.84
2.84
1.57
1.57
1.57
1.57
1.25
6 (41)
1.25
6.15
(42)
1.25
6 (41)
1.25
4.82
(33)
4.74
(33)
5.61
(39)
8.43
(58)
5.65
5.65
5.65
5.65
Table 4 Column properties and calculated plastic hinge length values (continued)
Kovacic
(1995)
Lehman
and
Moehle
(2000)
Calderone
(2000)
Dodd
(2000)
Moyer
and
Kowalsky
(2002)
d or
h, in.
(mm)
(1)
6.46
(164)
6.46
(164)
5.73
(146)
5.73
(146)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
7.9
(201)
7. 9
(201)
7.9
(201)
7. 9
(201)
7.9
(201)
7. 9
(201)
7.9
(201)
7. 9
(201)
7.9
(201)
18
(457)
18
(457)
Measured
Lp, in.
(mm)
(2)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
2.76 (70)
0.47
0.08
8.46
4.13
(105)
0.70
0.17
8.46
3.23 (82)
0.55
0.04
8.46
2.76 (70)
0.47
0.08
8.46
14 (356)
0.58
0.07
4.00
22 (559)
0.92
0.07
8.00
25 (635)
1.04
0.07
10.00
14 (356)
0.58
0.07
4.00
14 (356)
0.58
0.07
4.00
0.56
0.09
3.00
1.02
0.09
8.00
29 (737)
1.21
0.09
10.00
3.54 (90)
0.45
0.04
4.00
3.35 (85)
0.42
0.05
4.00
0.62
0.05
7.00
0.67
0.05
7.00
0.97
0.04
10.00
0.75
0.06
10.00
3.35 (85)
0.42
0.04
4.00
5.31
(135)
0.67
0.04
7.00
3.74 (95)
0.47
0.04
4.00
0.50
0.04
4.67
0.50
0.04
4.67
13.5
(343)
24.5
(622)
4.92
(125)
5.31
(135)
7.68
(195)
5.91
(150)
8.97
(228)
8.97
(228)
225
L, in.
(mm)
(6)
50
(1270)
50
(1270)
50
(1270)
50
(1270)
96
(2438)
192
(4877)
240
(6096)
96
(2438)
96
(2438)
72
(1829)
192
(4877)
240
(6096)
31.6
(803)
31.6
(803)
55.3
(1405)
55.3
(1405)
79
(2007)
79
(2007)
31.6
(803)
55.3
(1405)
31.6
(803)
84
(2134)
84
(2134)
db,
in.
(mm)
(7)
0.47
(12)
0.47
(12)
0.47
(12)
0.47
(12)
0.63
(16)
0.63
(16)
0.63
(16)
0.63
(16)
0.63
(16)
0.75
(19)
0.75
(19)
0.75
(19)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.75
(19)
0.75
(19)
fy, ksi
(MPa)
(8)
58
(400)
58
(400)
58
(400)
58
(400)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
60
(414)
60
(414)
f'c,
ksi
(MPa)
(9)
9.33
(64)
9.4
(65)
9.11
(63)
9.46
(65)
4.2
(29)
4.2
(29)
4.2
(29)
4.2
(29)
4.2
(29)
l
(%)
(10)
5.65
5.65
4.71
4.71
1.50
1.50
1.50
0.75
3.00
4 (28)
2.70
4 (28)
2.70
4 (28)
2.70
5.9
(41)
4.64
(32)
5 (34)
4.39
(30)
5.42
(37)
4.22
(29)
5.99
(41)
5.86
(40)
5.35
(37)
4.74
(33)
4.96
(34)
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
2.00
2.00
Table 4 Column properties and calculated plastic hinge length values (continued)
d or Measured
h, in.
Lp, in.
(mm)
(mm)
Lp/d P/f'cAg
(1)
(2)
(3)
(4)
18
8.97
Moyer
0.50
0.04
(457)
(228)
and
Kowalsky
18
8.97
0.50
0.04
(2002)
(457)
(228)
15.95
22.01
1.38
0.08
(405)
(559)
Phan
15.93
17.05
1.07
0.08
(2007)
(405)
(433)
16.05
17.01
1.06
0.08
(408)
(432)
1: Section dimension of the column tested.
L/d
(5)
4.67
4.67
4.50
4.50
4.50
L, in.
(mm)
(6)
84
(2134)
84
(2134)
72
(1829)
72
(1829)
72
(1829)
db,
in.
(mm)
(7)
0.75
(19)
0.75
(19)
0.75
(19)
0.75
(19)
0.75
(19)
fy, ksi
(MPa)
(8)
60
(414)
60
(414)
73.6
(507)
73.6
(507)
73.6
(507)
f'c,
ksi
(MPa)
(9)
4.6
(32)
4.92
(34)
4.7
(32)
4.72
(33)
4.81
(33)
l
(%)
(10)
2.00
2.00
2.00
2.20
2.00
2: Measured plastic hinge length displacement during testing.

3: The ratio of plastic hinge length to the section dimension
5: Aspect ratio
6: The length of the columns
7: The diameter of longitudinal reinforcement
8: Yield strength of the longitudinal reinforcement
9: Compressive strength of the concrete
10: The ratio of longitudinal reinforcement
11: The average plastic hinge length result was used from the experimental data ranging from 12 in. to 14
in. as given in PEER report.
Table 5 Column properties and measured plastic hinge length values
Cheok
and
Stone
(1986)
d or
h, in.
(mm)
(1)
9.8
(249)
9.8
(249)
9.8
(249)
9.8
(249)
9.8
(249)
Measured
Lp, in.
(mm)
(2)
11.6
(295)
10.2
(259)
7.5
(191)
9.6
(244)
10.6
(269)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
1.18
0.20
3.00
1.05
0.10
6.00
0.76
0.10
3.00
0.98
0.20
3.00
1.08
0.10
6.00
226
L, in.
(mm)
(6)
29.5
(749)
59
(1499)
29.5
(749)
29.5
(749)
59
(1499)
db,
in.
(mm)
(7)
1.75
(44)
1.75
(44)
1.75
(44)
1.75
(44)
1.75
(44)
fy, ksi
(MPa)
(8)
57
(393)
57
(393)
57
(393)
57
(393)
57
(393)
f'c,
ksi
(MPa)
(9)
3.35
(23)
3.68
(25)
3.55
(24)
3.53
(24)
3.37
(23)
l
(%)
(10)
1.45
0.70
1.45
1.45
0.70
Table 5 Column properties and measured plastic hinge length values (continued)
Wong et
al.
(1990)
Kunnath
(1997)
Bae
(2005)
Watson
(1994)
Hachem
(2003)
(11)
d or
h, in.
(mm)
(1)
Measured
Lp, in.
(mm)
(2)
15.75
(400)
9.84
(250)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
24
(610)
24
(610)
15.75
(400)
15.75
(400)
7.09
(180)
7.12
(181)
5.91
(150)
5.90
(150)
7.87
(200)
7.68
(195)
8.66
(220)
7.87
(200)
11.76
(299)
11.28
(287)
4.09
(104)
5.2
(132)
15.75
(400)
L, in.
(mm)
(6)
db,
in.
(mm)
(7)
fy, ksi
(MPa)
(8)
f'c,
ksi
(MPa)
(9)
l
(%)
(10)
31.5
(800)
0.63
(16)
61.35
(423)
5.51
(38)
3.20
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
103.5
(2629)
103.5
(2629)
63
(1600)
63
(1600)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.63
(16)
0.63
(16)
0.63
(16)
0.63
(16)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
64.67
(446)
64.67
(446)
64.67
(446)
68.73
(474)
4.21
(29)
4.21
(29)
4.76
(33)
4.76
(33)
4.71
(32)
3.92
(27)
3.92
(27)
3.92
(27)
6.38
(44)
6.38
(44)
5.8
(40)
5.95
(41)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
1.42
0.19
2.00
0.59
0.09
4.50
0.59
0.09
4.50
0.49
0.09
4.50
0.49
0.09
4.50
0.66
0.09
4.50
0.64
0.10
4.50
0.72
0.10
4.50
0.66
0.10
4.50
0.49
0.20
4.31
0.47
0.20
4.31
0.26
0.10
4.00
0.33
0.30
4.00
6.46
(164)
0.41
0.30
4.00
63
(1600)
0.63
(16)
68.73
(474)
5.8
(40)
1.51
15.75
(400)
4.09
(104)
0.26
0.30
4.00
63
(1600)
0.63
(16)
68.73
(474)
6.09
(42)
1.51
16
(406)
13
(330)
0.81
0.06
6.00
96
(2438)
0.5
(13)
60
(414)
5.7
(39)
1.20
227
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
1.25
1.25
1.51
1.51
Table 6 Maximum drift estimates
Authors
Test data
Park (1982)
Mander (1983)
Park (1987)
Priestley (1992)
Bae (2005)
Berry (2008)
Eq. 1
Eq. 2
max (in.)
(transverse
dir.)
5.3
5.9
9.4
9.2
10.4
4.0
7.1
6.5
7.1
lp (in.)
4.8
8.2
8.0
9.4
3.0
5.9
5.4
5.9
(% difference)
12.0
77.0
73.0
97.0
-25.0
35.0
23.0
23.0
282 in. [7.16 m]
348 in. [8.84 m]
348 in. [8.84 m]
282 in. [7.16 m]

14 in. [0.36 m]
Test Column
SHAKE TABLE
Abutment 1
SHAKE TABLE
Bent 1
SHAKE TABLE
Bent 2
Bent 3
Abutment 2
Alemdar Eq. 1
Bae
Park 82
5
Lp/d -measured/Lp/d-estimated
Park 87
Bae
4
Paulay 92
Berry
Mander 83
Berry 08
3
Park_82
2
Mander
1
Park_87
Alemdar Eq.1
0
0
100
200
300
400
500
fy *db/fc
600
700
Paulay
800
900
1000
Figure 2 Comparisons of the plastic hinge length expressions results with Equation 1
228
Alemdar Eq.
1
Bae
6
.
Park 82
5
4
Park 87
Berry
Park_82
Bae
Paulay 92
Mander 83
Berry 08
Mander
2
Park_87
1
Alemdar Eq. 1
Paulay
0
50
100
150
200
250
300
L (in.)
Figure 3 Comparisons of the plastic hinge length expressions results with Equation 1
Alemdar Eq. 2
Bae
Park_82
Park 82
Bae
Park 87
Berry
Paulay 92
4
Mander 83
Berry 08
3
2
Alemdar Eq. 2
1
0
0
200 Park_87
400
600 Paulay
800
Park_87
1000
fy *db/fc
1200
1400
1600
1800
Mander
2000
Figure 4 Comparisons of the plastic hinge length expressions results with Eq. 2
229
Alemdar Eq. 2
Bae
Park_82
Park 82
Berry
Park 87
Bae
Paulay 92
Mander 83
Berry 08
1
0
0 Eq. 2
Alemdar
20
40
60
Park_87
80
100
120
Paulay
L (in.)
Mander
Figure 5 Comparisons of the plastic hinge length expressions results with Eq. 2
Alemdar Eq. 1
4.5
Bae
Park 82
3.5
Park 87
Alemdar Eq. 1
2.5
Paulay 92
Bae
Mander 83
Berry 08
1.5
1
0.5
0
0
Alemdar Eq. 1
0.1
0.2
0.3
0.4
0.5
Bae
0.6
0.7
0.8
'
P/fc Ag (%)
Figure 6 Comparisons of the normalized plastic hinge length results with axial load level.
230
REFERENCES
1. ABAQUS, Version 6.8-2, Simulia, 2009, http://www.simulia.com.

2. Baker, A.L.L. (1956), Ultimate Load Theory Applied to the Design of
Reinforced and Prestressed Concrete Frames, Concrete Publications Ltd.,
London.
3. Baker, A.L.L. and Amarakone, A.M.N. (1964), Inelastic Hyperstatic Frame
Analysis, Proceedings International Symposium on the Flexural Mechanics of
Reinforced Concrete, ACI SP-12, Miami, pp. 85-142.
4. Mattock, A.H. (1964), Rotational Capacity of Hinging Regions in Reinforced
Concrete Beams, Proceedings International Symposium on the Flexural
Mechanics of Reinforced Concrete, ACI SP-12, Miami, pp. 143-181.
5. Mattock, A.H. (1967), Discussion of Rotational Capacity of Hinging Regions in
Reinforced Concrete Beams, by Corley, W.G., Journal of the Structural
Division, ASCE, V. 93, No. ST2, Apr., pp. 519-522.
6. Corley, W.G. (1966), Rotational Capacity of Reinforced Concrete Beams,
Journal of the Structural Division, ASCE, V. 92, No. ST5, Oct., pp. 121-146.
7. Park R., Priestley M. J. N., and Gill W. D. (1982), Ductility of Square-Confined
Concrete Columns, Struct. J., ASCE, 108, No. ST4, 929-950.
8. Mander J. B. (1983), Seismic Design of Bridge Piers, PhD Thesis, University
of Canterbury, Christ Church, New Zealand.
9. Soesianawati, M.T., Park, R. and Priestley, M.J.N. (1986), Limited Ductility
Design of Reinforced Concrete Columns, Research Report 86-10, Department of
231
Civil Engineering, University of Canterbury, Christchurch, New Zealand, Mar.,

208 pp.
10. Priestley, M. J. N., and Park R. (1987), Strength and Ductility of Concrete
Bridge Columns Under Seismic Loading, ACI Structural Journal, V.84, Issue 1,
pp. 61-76.
11. Paulay T., and Priestley M. J. N. (1992), Seismic Design of Reinforced Concrete
and Masonry Structures, John Wiley & Sons, Inc.
12. Sheikh, S.A. and Khoury, S.S. (1993), Confined Concrete Columns with Stubs,
ACI Structural Journal, V. 90, No. 4, July-Aug., pp. 414-431.
13. Sheikh, S.A., Shah, D.V. and Khoury, S.S. (1994), Confinement of HighStrength Concrete Columns, ACI Structural Journal, V. 91, No. 1, Jan.-Feb., pp.
100-111.
14. Bayrak, O. and Sheikh, S.A. (1997), High-Strength Concrete Columns under
Simulated Earthquake Loading, ACI Structural Journal, V. 94, No. 6, Nov.Dec., pp. 708-722.
15. Bae, S. (2005), Seismic Performance of Full-Scale Reinforced Concrete
Columns, PhD Dissertation, the University of Texas at Austin, Austin, TX, 311
pp.
16. Watson, S. and Park, R. (1994), Simulated Seismic Load Tests on Reinforced
Concrete Columns, Journal of Structural Engineering, ASCE, V. 120, No. 6,
June, pp. 1825-1849.
17. Phan V., Saiidi M.S., Anderson J., and Ghasemi H. (2007), Near-Fault Ground
Motion Effects on Reinforced Concrete Bridge Columns, Journal of Structural
Eng., V.133, Issue 7, pp. 982-989.
232
18. Berry, M., Lehman D. E., and Lowes L. N. (2008), Lumped-Plasticity Models
for Performance Simulation of Bridge Columns, ACI Structural Journal, V. 105,
No. 3, May-June, pp. 270-279.
19. Hachem M. M., Mahin, S.A., and Moehle J.P. Performance of Circular
Reinforced Concrete Bridge Columns under Bidirectional Earthquake Loading.
Report No. PEER 2003/06, Pacific Earthquake Engineering Research Center,
University of California at Berkeley, Feb. 2003, 490 pp.
20. Dodd L.L., The Dynamic Behavior of Reinforced Concrete Bridge Piers
Subjected to New Zealand Seismicity, PhD. Thesis, the University of
Canterbury Christchurch New Zealand, Feb. 1992, 448 pages.
21. Dodd, L. L., and Cooke, N., Capacity of Circular Bridge Columns Subjected to
Base Excitation, ACI Structural Journal, Vol. 97, No.2, March-April 2000,
pp.297-308.
22. Kunnath, Sashi, K.; El-Bahy, Ashraf; Taylor, Andrew; and Stone, William,
Cumulative Seismic Damage of Reinforced Concrete Bridge Piers, Technical
Report NCEER-97-0006, National Center for Earthquake Engineering Research,
September 1997, 147 pp.
23. Cheok, G.S.; and Stone, William C., Behavior of 1/6-Scale Model Bridge
Columns Subjected to Cycle Inelastic Loading, NBSIR 86-3494, Center for
Building Technology, National Engineering Laboratory, National Institute of
Standards and Technology, Gaithersburg.
24. Stone, William C.; and Cheok, Geraldine S., Inelastic Behavior of Full-Scale
Bridge Columns Subjected to Cyclic Loading, NIST BSS 166, Building Science
233
Series, Center for Building Technology, National Engineering Laboratory,

National Institute of Standard, 1989.
25. Wong, Y.L.; Paulay, T.; and Priestley, M.J.N., Squat Circular Bridge Piers Under
Multi-Directional Seismic Attack, Report 90-4, Department of Civil Engineering,
University of Canterbury, Christchurch, New Zealand, October 1990, 264 pp.
26. Tanaka, H. and Park, R. (1990), Effect of Lateral Confining Reinforcement on
the Ductile Behavior of Reinforced Concrete Columns, Research Report 90-2,
Department of Civil Engineering, University of Canterbury, Christchurch, New
Zealand, June, 458 pp.
27. Moyer, M.J. and Kowalsky, M.J., "Influence of tension strain on buckling of
reinforcement in RC bridge columns." ACI Structural Journal, 2002.
28. Calderone, A.J., Lehman, D.E.; Moehle, J.P.; "Behavior of Reinforced Concrete
Bridge Columns Having Varying Aspect Ratios and Varying Lengths of
Confinement," Pacific Earthquake Engineering Research Center, PEER 2000/08,
2000.
29. Lehman, D.E.; Moehle, J.P.; "Seismic Performance of Well-Confined Concrete
Bridge Columns," Pacific Earthquake Engineering Research Center, PEER
1998/01, Dec. 2000.
30. Firat Alemdar, Z.; Browning J.; Olafsen J.; "Photogrammetric Measurements of
RC Bridge Column Deformations, Engineering Structures, 2010 (submitted).
31. Firat Alemdar, Z.; Plastic hinging behavior of reinforced concrete bridge
columns Ph.D. dissertation, Lawrence (KS): the University of Kansas; 2010.
32. Kovacic D. (1995), Design of High-Strength Concrete Columns, M.Eng.
Thesis, the University of Melbourne.
234
CONCLUSIONS
This study focused on the plastic hinging behavior of reinforced concrete bridge
columns. A review of the literature revealed that most of the tests that have explored this
behavior have developed calibrated expressions to estimate regions of lumped plasticity,
which are then used to estimate maximum drift. There are a fewer number of tests that
have actually measured the spread of plasticity. In addition, the number of tested
elements that are representative of bridge columns (having lower axial load, lower
reinforcement ratios, and reasonable concrete compressive strengths) is limited. This
study sought to better define the plastic hinging behavior of reinforced concrete bridge
columns by taking detailed measurements of the column deformations in a large-scale 4span bridge test at the University of Nevada Reno, modeling the column using finite
elements, and developing separate expressions to calculate the maximum drift and the
spread of plasticity.
Three manuscripts were developed in the course of this research study. The
conclusions from each manuscript are repeated below.
From Manuscript 1: Photogrammetric Measurements of Reinforced

Concrete Bridge Column Deformations
Application of a simple photogrammetry method was used to evaluate the

deformations of a bridge column that was part of a bridge system tested at UNR.
According to the FDE index calculations, the lateral deformations of the bridge
column reduced from the photogrammetric data gave very good correlations with
235
the FE model of the column as well as the traditional instrumentation

measurements.
The rotation of a vertical line on the grid surface provided a good representation
of the column rotation.
The deformed shape of the bridge column was constructed using the lateral
displacements of the points on the bottom and top grid surfaces and matched
well with the images at the time of maximum drift of Test 4D.
From Manuscript 2: Modeling surface Deformations and Hinging Regions in

Reinforced Concrete Bridge Columns
Using a 1 in. (25 mm) mesh along the hinging regions of the bridge column
provided a similar force-displacement curve with the Modified Compression
Field Theory.
According to the FDE index analysis, using the yield strength of 80 ksi (552
MPa) gave the best correlations between the FE model and the traditional
instrumentations.
The flexural stiffness of the cap beam can be modeled by defining a rotational
spring with 5*10^8 lb-in. (56.5*10^6 N-m) capacity.
The tension and compression recovery variables were utilized to consider the
reduction in stiffness of the concrete.
The best FE model of the bridge column under dynamic loading was defined
using a stress-strain relationship for the longitudinal steel and concrete with the
dynamic magnification factors and including the tension stiffening effect.
236
The bond-slip effect was defined in the FE model to consider the slip of the
longitudinal steel at the column interfaces.
The FE model of the bridge column followed a similar deformed shape obtained
from the photogrammetry data.
From Manuscript 3: Plastic Hinge Length Expression for Reinforced

Concrete Bridge Columns
Two different plastic hinge length expressions were developed using the plastic
hinge length test data available in the literature. Equation 1 was produced based
on the plastic hinge length values that were calibrated to calculate the correct
maximum drift at the column. Equation 2 was developed using the measured
plastic hinge length values based on the spread of plastic curvature along the
height of the column.
The proposed equations gave the best fit at low axial loads.
The maximum drift capacity of the bridge column was estimated using the
proposed equations. The results show that the proposed equations give very good
estimates of the maximum drift capacity for the bridge column under dynamic
loading.
The plastic hinge length calculated using Eq. 1 gives shorter in value than that of
Eq. 2. More parametric analysis is required to evaluate the effect of the spread of
plasticity.
The study of plastic hinging regions for reinforced concrete bridge columns
shows that best estimates of the plastic hinge length depends on the goal of the analysis:
237
either for estimating the maximum drift capacity, or the spread of plasticity (>y). In
either case, it was found that plastic deformations can be measured with a simple
photogrammetry method during dynamic tests. The plastic behavior of the column was
also adequately represented using finite element modeling. Separate equations were
developed to represent the plastic hinge length for a lumped plasticity model (to estimate
the maximum drift) and for the spread of plasticity. The developed expressions provide
better estimates of plastic hinge lengths for lumped plasticity models at low axial loads
than any other expression that was considered. The proposed expression for a lumped
plasticity model was used effectively to estimate the maximum drift capacity. More
parametric analysis is needed, however, to understand the behavior of the spread of
plasticity.
238
REFERENCES FOR CHAPTERS 1 AND 2
ACI-ASCE Committee 428 (1968). Progress Report on Code Clauses for Limit
Design, ACI Journal, 65, No.9, pp. 713-715.
ACI Committee 318 (1989). Building Code Requirements for Reinforced Concrete and
Commentary (ACI 318-89/ACI 318R-89), American Concrete Institute, Detroit,
Michigan, 351 pp.
ACI Committee 318 (2005). Building Code Requirements for Reinforced Concrete and
Commentary (ACI 318-05/ACI 318R-05), American Concrete Institute, Detroit,
Michigan, 423 pp.
Ang Beng Ghee, Priestley M. J. N., and Park R. (1981). Ductility of Reinforced
Concrete Bridge Piers under Seismic Loading, Department of Civil Engineering
Research Report 81-3, University of Canterbury.
Bae, S. (2005). Seismic Performance of Full-Scale Reinforced Concrete Columns,

PhD Dissertation, the University of Texas at Austin, Austin, TX, 311 pp.
Bae, S. and Bayrak O. Plastic Hinge Length of Reinforced Concrete Columns, ACI
Structural Journal, V. 105, N. 3, May-June 2008, pp. 290-300.
239
Baker, A.L.L. (1956). Ultimate Load Theory Applied to the Design of Reinforced and
Prestressed Concrete Frames, Concrete Publications Ltd., London.
Baker, A.L.L. and Amarakone, A.M.N. (1964). Inelastic Hyperstatic Frame Analysis,
Proceedings International Symposium on the Flexural Mechanics of Reinforced
Concrete, ACI SP-12, Miami, pp. 85-142.
Bayrak, O. and Sheikh, S.A. (1997). High-Strength Concrete Columns under Simulated
Earthquake Loading, ACI Structural Journal, V. 94, No. 6, Nov.-Dec., pp. 708-722.
Bayrak, O. (1999). Seismic Performance of Rectilinearly Confined High Strength

Concrete Columns, Ph.D. dissertation, University of Toronto, Ontario, Canada, 339 pp.
Berry, M., Parish, M., and Eberhard, M. (2004). PEER Structural Performance
Database Users Manual, PEER Research Report, University of California-Berkeley,
Berkeley, CA, 2004, http://nisee.berkeley.edu/spd.
Berry, M., Lehman D. E., and Lowes L. N. (2008). Lumped-Plasticity Models for
Performance Simulation of Bridge Columns, ACI Structural Journal, V. 105, No. 3,
May-June, pp. 270-279.
Chan W. W.L. (1954). An Investigation of the Characteristics of Plastic Hinges in

Reinforced Concrete, PhD. Thesis, the University of London.
240
Chan W. W. L. (1955). The Ultimate Strength and Deformation of Plastic Hinges in

Reinforced Concrete Frameworks, Magazine of Concrete Research, Vol. 7, No. 21, pp.
121-132.
Chan W. W. L. (1962). Rotation of Reinforced Concrete Plastic Hinges at Ultimate

Load, Magazine of Concrete Research, V. 14, Issue 41, pp. 63-72.
Corley, W.G. (1966). Rotational Capacity of Reinforced Concrete Beams, Journal of

the Structural Division, ASCE, V. 92, No. ST5, Oct., pp. 121-146.
Davey B. E., and Park R. (1975). Reinforced Concrete Bridge Piers Under Seismic
Loading, Department of Civil Engineering Research Report 75-3, University of
Canterbury.
Dodd L.L. (1992). The Dynamic Behavior of Reinforced Concrete Bridge Piers
Subjected to New Zealand Seismicity, PhD. Thesis, the University of Canterbury
Christchurch New Zealand, 448 pages.
Dodd, L. L., and Cooke, N. (2000). Capacity of Circular Bridge Columns Subjected to
Base Excitation, ACI Structural Journal, Vol. 97, No.2, pp.297-308.
Ernst G. C. (1957). Plastic Hinging at the Intersection of Beams and Columns, Journal
of American Concrete Institute, Vol. 28, No.12, pp.1119-1144.
241
Gill W. D., Park R., and Priestley M. J. N. (1979). Ductility of Rectangular Reinforced
Concrete Columns with Axial Load, Department of Civil Engineering Research Report
79-1, University of Canterbury.
Hachem M. M., Mahin, S.A., and Moehle J.P. (2003). Performance of Circular
Reinforced Concrete Bridge Columns under Bidirectional Earthquake Loading, Report
No. PEER 2003/06, Pacific Earthquake Engineering Research Center, University of
California at Berkeley, 490 pp.
I.C.E. Research Committee (1962). Ultimate Load Design of Concrete Structures,

Proceedings of I.C.E. (London), Vol. 21, pp. 399-442.
Kovacic D. (1995). Design of High-Strength Concrete Columns, M.Eng. Thesis, the

University of Melbourne.
Mander J. B. (1983). Seismic Design of Bridge Piers, PhD Thesis, University of

Canterbury, Christ Church, New Zealand.
Mander J. B., Priestley M. J. N., and Park R. (1984). Seismic Design of Bridge Piers,
Department of Civil Engineering Research Report 84-2, University of Canterbury, 483
pp.
242
Mattock, A.H. (1964). Rotational Capacity of Hinging Regions in Reinforced Concrete

Beams, Proceedings International Symposium on the Flexural Mechanics of Reinforced
Concrete, ACI SP-12, Miami, pp. 143-181.
Mattock, A.H. (1967). Discussion of Rotational Capacity of Hinging Regions in

Reinforced Concrete Beams, by Corley, W.G., Journal of the Structural Division,
ASCE, V. 93, No. ST2, Apr., pp. 519-522.
McCollister H. M., Siess C. P., and Newmark N. M. (1954). Load Deformation

Characteristic of Simulated Beam-Column Connections in Reinforced Concrete,
University of Illinois, Civil Engineering Studies, Structural Research Series.
Mendis, P. (2001). Plastic Hinge Lengths of Normal and High-Strength Concrete in

Flexure, Advances in Structural Engineering, V. 4, No. 4, Oct., pp. 189-195.
Munro I. R. M., Priestley M. J. N., and Park R. (1976). Seismic Behavior of Reinforced
Concrete Bridge Piers, Department of Civil Engineering Research Report 76-7,
University of Canterbury.
Ng Kit Heng, Priestley M. J. N., and Park R. (1978). Seismic Behavior of Circular
Reinforced Concrete Bridge Piers, Department of Civil Engineering Research Report
78-14, University of Canterbury.
243
Park R., Priestley M. J. N., and Gill W. D. (1982). Ductility of Square-Confined

Concrete Columns, Struct. J., ASCE, 108, No. ST4, 929-950.
Paulay T., and Priestley M. J. N. (1992). Seismic Design of Reinforced Concrete and
Masonry Structures, John Wiley & Sons, Inc.
Phan V., Saiidi M.S., Anderson J., and Ghasemi H. (2007). Near-Fault Ground Motion
Effects on Reinforced Concrete Bridge Columns, Journal of Structural Eng., V.133,
Issue 7, pp. 982-989.
Poologasoundranayagam K. (1960). An analytical and experimental investigation of the

formation and behavior of plastic hinges in prestressed and reinforced concrete frames,
Ph.D. Thesis, pp. 187.
Potangaroa R. T., Priestley M. J. N., and Park R. (1979). Ductility of Spirally-Confined

Reinforced Concrete Columns under Seismic Loading, Department of Civil
Engineering Research Report 79-8, University of Canterbury.
Priestley M. J. N., Park R., and Potangaroa R. T. (1981). Ductility of Spirally-Confined

Reinforced Concrete Columns, Proceedings, ASCE, V.107, ST1, pp.181-202.
Priestley, M. J. N., and Park R. (1987). Strength and Ductility of Concrete Bridge
Columns under Seismic Loading, ACI Structural Journal, V.84, Issue 1, pp. 61-76.
244
Restrepo J. I., Seible F., Stephan B., and Schoettler J. M. (2006). Seismic Testing of
Bridge Columns Incorporating High-Performance Materials, ACI Structural Journal,
V.103, Issue 4, pp. 496-504.
RRU Project 4805 (1983). Ductility of Reinforced Concrete Bridge Piers Under
Seismic Loading, Road Research Unit Newsletter, RRU NL 75, National Roads Board,
Wellington, New Zealand, pp.27-28.
Sakai, K. and Sheikh, S.A. (1989). What Do We Know about Confinement in

Reinforced Concrete Columns? (A Critical Review of Previous Work and Code
Provisions), ACI Structural Journal, V. 86, No. 2, Mar.-Apr., pp. 192-207.
Sheikh, S.A. and Khoury, S.S. (1993). Confined Concrete Columns with Stubs, ACI
Structural Journal, V. 90, No. 4, July-Aug., pp. 414-431.
Sheikh, S.A., Shah, D.V. and Khoury, S.S. (1994). Confinement of High-Strength
Concrete Columns, ACI Structural Journal, V. 91, No. 1, Jan.-Feb., pp. 100-111.
Soesianawati, M.T., Park, R. and Priestley, M.J.N. (1986). Limited Ductility Design of
Reinforced Concrete Columns, Research Report 86-10, Department of Civil
Engineering, University of Canterbury, Christchurch, New Zealand, Mar., 208 pp.
Standards Association of New Zealand (1982). New Zealand Standard Code of Practice
for the Design of Concrete Structures, NZS 3101, Wellington, New Zealand.
245
Watson, S. and Park, R. (1994). Simulated Seismic Load Tests on Reinforced Concrete
Columns, Journal of Structural Engineering, ASCE, V. 120, No. 6, June, pp. 18251849.
Tanaka, H. and Park, R. (1990). Effect of Lateral Confining Reinforcement on the

Ductile Behavior of Reinforced Concrete Columns, Research Report 90-2, Department
of Civil Engineering, University of Canterbury, Christchurch, New Zealand, June, 458
pp.
246
APPENDIX A
Appendix A contains the comparisons of rotations at each LVDT location for the
three different tests as described in Chapter 2.
Comparison (Test 2 Bottom)

0.015
Vertical_Rotation
Cross-sec. Rot.
0.01
Rotation (rad)
0.005
0
0
10
15
20
25
30
-0.005
-0.01
-0.015
Time (sec)
Fig. A-1: Comparison of rotations (local vs. average) at Point 3-8
247
35

0.015
Vertical_Rotation
Cross-sec. Rot.
0.01
Rotation (rad)
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
Time (sec)

Comparison (Test 2 Top)
0.03
Vertical_Rotation
Cross-sec. Rot.
0.02
Rotation (rad)
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec)
248
35

0.015
Vertical_Rotation
Cross-sec. Rot.
0.01
Rotation (rad)
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
Time (sec)

Comparison (Test 4D Bottom)
0.02
Cross-Sec. Rot.
Vertical Rot.
0.015
0.01
Rotation (rad)
0.005
0
-0.005
10
15
20
25
30
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec)
249
35
Comparison (Test 4D Bottom)

0.03
Vertical_Rotation
Cross-sec. Rot.
0.02
Rotation (rad)
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
-0.04
Time (sec)

Comparison (Test 4D Top)
0.05
Vertical_Rotation
Cross-sec. Rot.
0.04
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec)
250
35
Comparison (Test 4D Top)

0.05
Vertical_Rotation
0.04
Cross-sec. Rot.
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec)

0.05
Vertical_Rotation
Cross-sec. Rot.
0.04
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec)
251
35

0.06
Vertical_Rotation
Cross-sec. Rot.
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec)

0.08
Vertical_Rotation
Cross-sec. Rot.
0.06
Rotation (rad)
0.04
0.02
0
0
10
15
20
25
30
-0.02
-0.04
-0.06
Time (sec)
252
35

0.08
Vertical_Rotation
Cross-sec. Rot.
0.06
Rotation (rad)
0.04
0.02
0
0
10
15
20
25
30
-0.02
-0.04
-0.06
Time (sec)
253
35
APPENDIX B
Appendix B presents the comparisons of lateral displacements at each point on

the grid surface for the three different tests described in Chapter 2.
Note: Comparisons are done with the combined movements of DT7, DS5 and DS1
instrumentation. Note that there is difference of amplitude between the results because of
a difference in height at the points of measurements.
0.4
Photogrammetry
0.3
components
0.2
Displacement (in.)
0.1
0
0
10
15
20
25
30
35
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
Time (sec)
Fig. B-1: Lateral movement of Point 2 for Test 2 (1 in. = 254 mm)
254
40
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
255
40
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
256
40
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
257
40
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
258
40
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
259
40
0.4
Photogrammetry
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
0.8
Photogrammetry
0.6

components
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
260
40
0.8
Photogrammetry
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
0.8
Photogrammetry
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
261
40
0.8
Photogrammetry
0.6

components
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
0.8
Photogrammetry
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
262
40
0.8
Photogrammetry
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
0.8
Photogrammetry
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
263
40
1
Photogrammetry
0.8

components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
-1
Time (sec)
0.8
Photogrammetry
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
264
40
1
Photogrammetry
components
0.8
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
1
Photogrammetry
0.8

components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
265
40
0.8
Photogrammetry
0.6

components
0.4
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
-1
Time (sec)
1
Photogrammetry
0.8

components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
266
40
1.5
Photogrammetry
components
Displacement (in.)
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
-2
Time (sec)
Fig. B-26: Lateral movement of Point 2 for Test 4D (1 in. = 254 mm)
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
-0.5
-1.0
-1.5
-2.0
Time (sec)
267
35
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
Time (sec)
268
40
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
Time (sec)
269
40
1.5
Photogrammetry
components
Displacement (in.)
0.5
0
0
10
15
20
25
30
35
-0.5
-1
-1.5
-2
Time (sec)
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
-0.5
-1.0
-1.5
-2.0
Time (sec)
270
35
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
1.5
Photogrammetry
components
Displacement (in.)
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
-2
Time (sec)
271
35
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
1.5
Photogrammetry
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
Time (sec)
272
40
2.5
Photogrammetry
2

components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
-0.5
-1
-1.5
Time (sec)
2.5
Photogrammetry
2
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
273
35
2.5
Photogrammetry
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
2.5
Photogrammetry
2

components
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
274
35
2.5
Photogrammetry
components
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
3
Photogrammetry
2.5

components
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
275
2.5
Photogrammetry
2

components
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
2.5
Photogrammetry
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
276
35
2.5
Photogrammetry
2

components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
-0.5
-1
-1.5
Time (sec)
2.5
Photogrammetry
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
277
35
2.5
Photogrammetry
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
Time (sec)
2.5
Photogrammetry
2

components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
278
35
2.0
Photogrammetry
1.5

components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
2.0
Photogrammetry
1.5

components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
279
40
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
2.0
Photogrammetry
1.5

components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
280
40
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
281
40
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
282
40
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
283
40
2.0
Photogrammetry
1.5

components
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
2.0
Photogrammetry
1.5
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
284
40
5.00
Photogrammetry
4.00

components
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
5.00
Photogrammetry
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
285
40
5.00
Photogrammetry
4.00

components
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
5.00
Photogrammetry
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
286
40
5.00
Photogrammetry
4.00

components
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
5.00
Photogrammetry
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
287
40
5.00
Photogrammetry
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
5.00
Photogrammetry
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
288
40
5.00
Photogrammetry
components
4.00
3.00
Displacement (in.)
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
5.00
Photogrammetry
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
289
40
5.0
Photogrammetry
4.0
DT7 & (DS5+DS1)

combined components
Displacement (in.)
3.0
2.0
1.0
0.0
0
10
15
20
25
30
35
40
-1.0
-2.0
-3.0
Time (sec)
5.00
Photogrammetry
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
290
40

Diseño Capacidad 6 PDF

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Diseño Capacidad 6 PDF

Caricato da

Copyright:

Formati disponibili

PLASTIC HINGING BEHAVIOR OF REINFORCED CONCRETE BRIDGE

Submitted to the graduate degree program in Civil

Date Defended: April 26, 2010

The Dissertation Committee for Zeynep Firat Alemdar certifies

PLASTIC HINGING BEHAVIOR OF REINFORCED CONCRETE BRIDGE

Date approved: _________________________

The location of inelastic deformations in reinforced concrete bridge columns has

I want to express my sincere gratitude to my advisors, Professors JoAnn

Zeynep Firat Alemdar

LIST OF TABLES FOR CHAPTERS 1 AND 2.................................................................. iv

DETERMINATION OF PLASTIC HINGE LENGTH ........................................ 1

LITERATURE REVIEW OF PLASTIC HINGE LENGTH (STATIC TESTS).. 3

The Institution of Civil Engineers Committee Report (1962) ................... 4

ACI Limits (1968) ..................................................................................... 8

Park, Priestley, and Gill (1982) ................................................................. 9

Mander (1983) ......................................................................................... 11

Priestley and Park (1987)......................................................................... 13

Sakai and Sheikh (1989).......................................................................... 17

Tanaka and Park (1990)........................................................................... 17

Paulay and Priestley (1992) ..................................................................... 19

Sheikh and Khoury (1993), Sheikh, Shah and Khoury (1994)................ 24

Kovacic (1995) ........................................................................................ 27

Bayrak and Sheikh (1997, 1999) ............................................................. 28

Bae (2005) ............................................................................................... 31

Restrepo, Seible, Stephan, and Schoettler (2006).................................... 40

Phan V., Saiidi M.S., Anderson J., and Ghasemi H. (2007).................... 45

Berry, Lehman, and Lowes (2008) .......................................................... 46

LITERATURE REVIEW OF PLASTIC HINGE LENGTH (DYNAMIC

Dodd et al. (2000) .................................................................................... 49

Hachem et al. (2003)................................................................................ 52

OBJECTIVE AND SCOPE................................................................................. 57

2 PHOTOGRAMMETRIC MEASUREMENTS OF CONCRETE COLUMN

LARGE SCALE FOUR-SPAN RC BRIDGE TEST .......................................... 67

Definition of Points on Surface ............................................................... 78

Rotations ................................................................................................ 122

3 MANUSCRIPT 1: PHOTOGRAMMETRIC MEASUREMENTS OF RC BRIDGE

LIST OF TABLES FOR CHAPTERS 1 AND 2

Table 1.2 Experimental and predicted plastic hinge lengths.16

Table 1.3 Details of Column Specimens (Tanaka and Park 1990)...19

LIST OF FIGURES FOR CHAPTERS 1 AND 2

Figure 1.5 Theoretical curvature relationships for a prismatic reinforced concrete

Figure 1.10 Effect of Amount of Longitudinal Reinforcement (Bae 2005) ......................38

Figure 1.13 Plasticity spread coefficient ((Restrepo et al. 2006) ...................................43

Fig. 2.1 Cross-section of the column .................................................................................60

Fig. 2.2 Grid Setup.............................................................................................................61

Fig. 2.3 Location of the column and the tower position....................................................62

Fig. 2.4 The aluminum tower setup ...................................................................................63

Fig. 2.5 Acceleration history of the Rinaldi earthquake record.........................................64

Fig. 2.7 Rotation @ Grid Level A vs. Time (Rinaldi 0.95g).............................................66

Fig. 2.8 Elevation view of the four-span bridge ................................................................68

Fig. 2.9 Elevation and side view of the Bent1 ...................................................................68

Fig. 2.10 Elevation and side view of the Bent2 .................................................................69

Fig. 2.11 Elevation and side view of the Bent 3 ................................................................69

Fig. 2.12 Grid systems on the Bent3 east column surface.................................................71

Fig. 2.14 The aluminum tower and four cameras ..............................................................72

Fig. 2.28 Pixel intensities of two different levels in an image ..........................................88

Fig. 2.31 Rotation of horizontal line using Points 7 and 9 ................................................90

Fig. 2.32 Average rotation of four Points on the column surface......................................91

Fig. 2.34 Corner (a) rotation of Point 12 and 13 ...............................................................93

Fig. 2.35 Corner (b) rotation of Point 12 and 13 ...............................................................93

Fig. 2.36 Corner (c) rotation of Point 12 and 13 ...............................................................94

Fig. 2.37 Corner (d) rotation of Point 12 and 13 ...............................................................94

Fig. 2.38 Average rotation of Point 12 and 13 ..................................................................95